# Buy more credits
Source: https://easy-peasy.ai/docs/account/buy-credits

Top up image & video credits or words when your plan allowance runs low, without changing your plan.

If you need more before your allowance resets, you can buy extra image & video credits, and on larger plans, extra words. What you buy is added to your account as **bonus credits** or **bonus words**.

<Note>
  Buying extra credits is for paid plans. On the Free plan, [upgrade](/docs/account/subscription#upgrade-or-change-your-plan) instead.
</Note>

## Buy image & video credits

<Steps>
  <Step title="Click Buy more credits">
    You'll find the button on [**Settings** > **Subscription**](https://easy-peasy.ai/settings/billing), on [**Settings** > **Usage**](https://easy-peasy.ai/settings/usage), and on the image and video pages.
  </Step>

  <Step title="Choose an option">
    **Purchase Options** lists one-time packs of 100 or 500 credits, and bigger plans you can upgrade to. A bigger plan usually costs less per credit, and the dialog shows the price per credit for each option.

    <Frame>
      <img alt="The Purchase Options dialog: plan upgrades with credits per month, and one-time purchases of 100 and 500 credits with a Buy Now button" />
    </Frame>
  </Step>

  <Step title="Pay">
    Click **Buy Now** and pay on the Stripe checkout page. You can change the quantity there to buy up to 10 packs at once.
  </Step>
</Steps>

The credits appear as **Bonus credits** on the Usage page as soon as the payment goes through.

## Buy more words

On Unlimited 200 and larger plans, [**Settings** > **Usage**](https://easy-peasy.ai/settings/usage) shows **Buy more words** next to **Words Usage**. Each purchase adds 100,000 bonus words.

## Buy for a team member

If you own a team, you can buy credits or words for one member from [**Settings** > **Team**](https://easy-peasy.ai/settings/team). They go to that member's balance, not yours.

## How bonus credits and words are used

* They're used only after your plan's monthly allowance runs out.
* They don't reset. Whatever you don't use carries over to the next period.
* They show up separately on the Usage page as **Bonus credits** and **Bonus words**.

## Get bonus credits by referring friends

Share your link from [**Settings** > **Refer a friend**](https://easy-peasy.ai/settings/referrals). When someone you invite becomes a paying customer, you both get 50,000 bonus words and 150 bonus credits. If they choose Unlimited or a larger plan, you both get 100,000 words and 300 credits instead.


# Data controls
Source: https://easy-peasy.ai/docs/account/data-controls

Bring your ChatGPT conversations into Marky Agent, and manage or delete the websites, apps, and domains Marky built for you.

Open **Settings** > **Data Controls**.

## Import your ChatGPT history

Move your ChatGPT conversations into Marky Agent, so you can search them and continue them with any model.

<Frame>
  <img alt="Import from ChatGPT: Bring your ChatGPT conversations into Marky Agent, with the instruction to request an export in ChatGPT under Settings, Data controls, Export data, and an upload area, Choose your ChatGPT export, ZIP from OpenAI, or conversations.json" />
</Frame>

<Steps>
  <Step title="Export your data from ChatGPT">
    In ChatGPT, go to **Settings** > **Data controls** > **Export data**. OpenAI emails you a link to a ZIP file, usually within a few minutes to a few hours. Download it.
  </Step>

  <Step title="Choose the file">
    Click **Import from ChatGPT**, then **Choose your ChatGPT export**, and pick the ZIP, up to 2 GB. The file is read in your browser, and you see how many conversations, messages, and images were found.
  </Step>

  <Step title="Choose the options">
    * **Also teach Marky about me**, on by default, adds what your ChatGPT history says about you, such as your job, projects, and preferences, to [Marky's memory](/docs/marky/memory).
    * **Replace conversations I already imported**: turn it on to import again conversations you imported before. Otherwise they're skipped.
  </Step>

  <Step title="Click Import">
    Keep the tab open until it finishes. Your conversations appear in Marky Agent in a **ChatGPT** folder, with their original dates.
  </Step>
</Steps>

Good to know:

* Images you uploaded to ChatGPT come along. Images ChatGPT generated aren't in OpenAI's export, so they can't be imported.
* Voice conversations are imported as text.
* Importing is free and doesn't use words. Continuing a conversation in Marky uses words as usual.
* Imported conversations are private to you, even on a Teams plan.

## Your websites, apps, and domains

**My Websites** and **My Apps** list what [Marky built](/docs/marky/websites) for you, with the link, whether it's **Public** or a **Draft**, and when it last changed. The ⋯ menu on each has:

| Option | What it does |
| - | - |
| **Settings** | Opens the site's settings: its address, custom domain, and more |
| **Open chat** | Opens the Marky chat it was built in |
| **Visit site** | Opens the published site |
| **GitHub** | Opens the linked GitHub repository, if there is one |
| **Delete** | Deletes the site or app |

<Warning>
  **Delete** takes the site offline and deletes its database and files. It can't be undone. The chat it was built in stays, and so do domains you bought for it. Only the owner can delete a site.
</Warning>

**My Domains** lists the domains you bought through Marky, with their status, registration and expiry dates, and whether they renew automatically.

<Columns>
  <Card title="Websites and apps" icon="globe" href="/docs/marky/websites">
    Build and publish with Marky.
  </Card>

  <Card title="Profile and preferences" icon="user" href="/docs/account/profile">
    Your details, or delete your account.
  </Card>
</Columns>


# HIPAA and the BAA
Source: https://easy-peasy.ai/docs/account/hipaa

Sign our Business Associate Addendum to use the clinical note tools with patient information, with zero retention of their results.

If you work with protected health information in the US, sign the Easy-Peasy.AI **Business Associate Addendum (BAA)** before using it in the clinical note tools. Signing designates your account as a HIPAA account and turns on zero retention for those tools.

The BAA is available on paid plans.

## Sign the BAA

<Steps>
  <Step title="Open the BAA">
    Go to [easy-peasy.ai/settings/baa](https://easy-peasy.ai/settings/baa), or click the HIPAA banner on a clinical note tool, such as the SOAP Note Generator.
  </Step>

  <Step title="Read it">
    Read the agreement in the page, or click **Download BAA** to save a PDF copy. It lists the services it covers.
  </Step>

  <Step title="Tick the three boxes">
    <Frame>
      <img alt="BAA Acceptance with three checkboxes: I acknowledge that I have read and understood this agreement, I take responsibility for the terms, and I agree to the Zero-Retention policy, which deletes content from the Psychotherapy Progress Note, SOAP Note, DAP Note, Treatment Plan, and BIRP Note Generators within 24 hours, and the Accept EASY-PEASY.AI BAA and Designate HIPAA Account button" />
    </Frame>
  </Step>

  <Step title="Click Accept EASY-PEASY.AI BAA and Designate HIPAA Account">
    A **HIPAA BAA** tab appears in **Settings**, with the date you accepted and a copy to download.
  </Step>
</Steps>

## Zero retention

Once you've signed, results from these tools are deleted automatically within 24 hours:

* Psychotherapy Progress Note Generator
* SOAP Note Generator
* DAP Note Generator
* Treatment Plan Generator
* BIRP Note Generator

Copy or save each note into your own records system as soon as it's written. After 24 hours it's gone from your History.

<Warning>
  Only use protected health information in the services the BAA covers. Other features, such as Marky Agent, AI Speech to Text, and agents, aren't part of it.
</Warning>

## End the BAA

On the **HIPAA BAA** tab, click **Terminate BAA Agreement** and confirm. Your account is no longer a HIPAA account, and new results from the note tools are kept like any others. You can sign the BAA again later on the same page.

<Columns>
  <Card title="Run a tool" icon="play" href="/docs/tools/run">
    How the note tools work.
  </Card>

  <Card title="Get help" icon="life-ring" href="https://easy-peasy.ai/support">
    Questions about compliance.
  </Card>
</Columns>


# Plans and credits
Source: https://easy-peasy.ai/docs/account/plans-and-credits

How Easy-Peasy.AI measures what you use, what the Free plan includes, and what happens when an allowance runs out.

Every plan comes with a monthly allowance, split into a few separate balances. Each feature draws from one of them, and you can watch all of them on [**Settings** > **Usage**](/docs/account/usage).

| Balance | What uses it | How it's counted |
| - | - | - |
| **Words** | Marky Agent, tools, and agents on some models | Words of text the AI writes, adjusted by the model's multiplier |
| **Image & video credits** | AI Images, AI Videos, talking videos, image editing, upscaling, headshots, interior design, and images or slides Marky makes | A price per image, or per second of video, set by each model |
| **Text-to-speech characters** | Text to speech, voice changer, dubbing, sound effects, the audio generator, and music | Characters of text you turn into speech; other audio features use a fixed amount per job |
| **Transcriptions** | AI Speech to Text and Record & Transcribe | One per file |
| **Agents** | Agents | How many agents you can have |

Compare what each paid plan includes on the [pricing page](https://easy-peasy.ai/pricing).

## Words

Marky Agent and the tools count the words the AI writes for you. Two things change the count:

* **The model's multiplier.** Every chat model has one, shown next to its name in Marky's model picker. A 0.5x model uses half a word per word it writes; a 2x model uses two. The default model is one of the cheapest.
* **What the model reads and thinks.** A model's reasoning counts as writing. Premium models such as Claude Opus and Claude Fable also charge for long conversations and large files they read, after a free allowance each time.

<Note>
  In the [Quickstart](/docs/get-started/quickstart), Claude Opus 5.5 Reasoning wrote a 110-word email and used 505 words, because its reasoning and the conversation it read count too. The same request on the default model would use a small fraction of that.
</Note>

On the Unlimited and Teams plans, some models are labelled **Unlimited** in the model picker and don't use words at all. In tools, the **Standard model** is also free of words on those plans.

A few more rules:

* On the Free plan, every chat model counts at 0.2x.
* On the Starter plan, models labelled **Unlimited** count at 0.1x.
* Uploading an image to a tool adds 150 words.
* Agents built on some models use your words when they reply to your customers.

## Image & video credits

Images, videos, and the other visual features share one pool of credits. What a job costs depends on the model, the resolution, and for videos, the length. The **Generate** button always shows the cost before you start, and videos are priced per second: a 5-second clip at 4 credits per second costs 20 credits.

* **Failed jobs are refunded automatically.** If a generation fails, its credits come back without you doing anything.
* **You can refund a result you don't like.** Click the thumbs-down on your own image or video and its credits are returned, up to 10% of your plan's monthly credits each billing period.
* **Deleting doesn't refund.** Once an image or video has been created, deleting it doesn't return its credits.

## Text-to-speech characters

Text to speech uses one character per character of your script, half that with **Turbo**, and one per ten characters with OpenAI voices. Other audio features take a set amount from the same balance:

| Feature | Uses |
| - | - |
| Voice changer and dubbing | About 2,000 per minute of audio |
| Sound effects and the audio generator | 1,000 per generation |
| Music and songs | 600 per generation (two tracks, 300 each) |

## When your allowance resets

* **Paid plans** reset every month on your billing date. Yearly plans reset monthly too, on the same day of the month you subscribed.
* **The Free plan** resets on the 1st of each month.

Unused allowance doesn't roll over. Credits and words you [buy separately](/docs/account/buy-credits) are different: they're used only after your plan allowance runs out, and whatever you don't use carries over to the next period.

## The Free plan

New accounts start on the Free plan. It's for trying things out:

| Balance | Free plan |
| - | - |
| Words | 2,000 per month |
| Image & video credits | 3 per month |
| Text-to-speech characters | 100 per month |
| Transcriptions | 2 per month |
| Agents | 1, answering up to 100 messages per 30 days |

On the Free plan:

* Tools use the **Standard model** and make one result at a time. Tools marked as paid can be tried once.
* Premium image and video models are locked, and videos are limited to a few models and short lengths.
* Image downloads are for paid plans. Free accounts can download while they still have bonus credits.

## What happens when you run out

| You run out of | What happens |
| - | - |
| Words | On Free and Starter, Marky and tools stop until your allowance resets. On Unlimited and Teams plans, you can keep using the **Unlimited** chat models and the **Standard model** in tools. |
| Image & video credits | You can't start new images or videos until your allowance resets. |
| Anything else | That feature pauses until your allowance resets. |

To keep going sooner, [buy more credits](/docs/account/buy-credits) or [change your plan](/docs/account/subscription).


# Profile and preferences
Source: https://easy-peasy.ai/docs/account/profile

Change your name, email address, password, dashboard language, and theme, or delete your account.

Open **Settings** in the sidebar. The **Profile** tab has your account details, and **Interface** has the language and theme.

<Frame>
  <img alt="Settings on the Profile tab: the tabs Profile, Team, Interface, Subscription, Usage, Data Controls, API/MCP, and Refer a friend, then First name and Last name with Save, Email address with Save, and Change Password" />
</Frame>

## Name and email

* **Name**: fill in **First name** and **Last name** and click **Save**.
* **Email**: type the new address and click **Save**. We send a confirmation link to the new address; the change takes effect once you click it.

## Password

Under **Change Password**, type a **New Password** and click **Update Password**. If you signed up with Google, this adds a password so you can also sign in with your email.

A password needs 12 to 128 characters, with an uppercase letter, a lowercase letter, a number, and a special character. It can't be a common password or contain your email address. The checklist under the field shows what's missing.

If you forget your password, click **Forgot your password?** on the sign-in page to get a reset link.

## Language and theme

On the **Interface** tab:

* **Language** changes the dashboard's language: English, Japanese, Spanish, French, Italian, German, Korean, Simplified Chinese, Thai, Arabic, Portuguese, Malay, or Vietnamese.
* **Theme** switches between light, dark, and your system's setting.

The globe and sun icons under the logo in the sidebar do the same.

Changing the dashboard language doesn't change what the AI writes. Pick the output language in each [tool](/docs/tools/run), or ask Marky in the language you want.

## Delete your account

<Frame>
  <img alt="Change Password with a New Password field and Update Password, and below it Delete Account, This action is irreversible, with a Delete Account button" />
</Frame>

<Steps>
  <Step title="Click Delete Account">
    At the bottom of the **Profile** tab.
  </Step>

  <Step title="Type DELETE to confirm">
    You can also tell us why you're leaving. Then your account is deleted.
  </Step>
</Steps>

<Warning>
  Deleting your account can't be undone. It cancels your subscriptions straight away, takes down the websites and apps Marky published for you, and deletes your chats, agents, custom voices, and everything you created. Download anything you want to keep first. Domains you bought aren't deleted.
</Warning>

To stop paying without deleting your account, [cancel your subscription](/docs/account/subscription) instead.

<Columns>
  <Card title="Manage your subscription" icon="credit-card" href="/docs/account/subscription">
    Change plan or cancel.
  </Card>

  <Card title="Data controls" icon="database" href="/docs/account/data-controls">
    Import from ChatGPT, and your sites and domains.
  </Card>
</Columns>


# Refer a friend
Source: https://easy-peasy.ai/docs/account/referrals

Share your invite link and get bonus words and credits, for you and your friend, when they subscribe.

Invite friends to Easy-Peasy.AI. When a friend you invited starts a paid plan, you both get bonus words and image & video credits.

| Your friend subscribes to | You each get |
| - | - |
| Unlimited or higher, a Teams plan, or Enterprise | 100,000 words and 300 credits |
| Any other paid plan, such as Starter or Unlimited 50 | 50,000 words and 150 credits |

Bonus words and credits don't expire, and they're used after your plan's monthly allowance. See [Buy more credits](/docs/account/buy-credits).

## Invite a friend

<Steps>
  <Step title="Copy your invite link">
    Open **Settings** > **Refer a friend** and click **Copy** next to **Copy your invite link**.

    <Frame>
      <img alt="Refer a friend: invite a friend and both of you receive 50,000 words and 150 image credits, or 100,000 words and 300 image credits when your friend subscribes to an Unlimited plan or higher, with the invite link and a Copy button, and an Affiliate Program link" />
    </Frame>
  </Step>

  <Step title="Share it">
    Send the link to friends or post it. They need to sign up through your link to count as your referral.
  </Step>

  <Step title="Get your bonus">
    The bonus is added to both accounts within 7 days after your friend starts a paid subscription.
  </Step>
</Steps>

At the top of the page, **Total referrals** counts everyone who signed up with your link, and **Subscribed referrals** those who subscribed. The table shows each one with its status: **Invited**, **Registered**, or **Subscribed**.

## Affiliate program

To earn money instead of credits, for example if you write about AI tools or run a community, join the affiliate program: **Affiliate Program** at the bottom of the page. Affiliates earn a 30% commission on the payments of the customers they refer, for as long as they stay subscribed.

<Columns>
  <Card title="Plans and credits" icon="coins" href="/docs/account/plans-and-credits">
    How words and credits work.
  </Card>

  <Card title="Buy more credits" icon="cart-shopping" href="/docs/account/buy-credits">
    How bonus balances are used.
  </Card>
</Columns>


# Manage your subscription
Source: https://easy-peasy.ai/docs/account/subscription

Upgrade or change your plan, update your card, download invoices, and cancel.

Everything about your plan is on [**Settings** > **Subscription**](https://easy-peasy.ai/settings/billing). You can also get there from the account menu under the logo.

<Frame>
  <img alt="The Subscription tab: the current plan with its renewal date, buttons for Change plan, Edit payment details, and View billing history, and the image and video credits used this period" />
</Frame>

The card at the top shows your plan, its status, and the date your allowance renews. Payments are handled by Stripe, so the buttons open Stripe's secure billing page in the same window.

<Note>
  On a team, only the account owner can manage the subscription. Team members see a message asking them to talk to the owner.
</Note>

## Upgrade or change your plan

<Steps>
  <Step title="Pick a plan">
    Scroll to **Choose your plan** on the same page. Switch between **Monthly** and **Yearly** at the top; yearly billing costs less per month. Click **Show all plans** to see the larger plans.
  </Step>

  <Step title="Click Upgrade">
    Stripe shows what you'll pay before you confirm. When you upgrade, you pay only the difference for the rest of the current period, and your billing date doesn't change.
  </Step>

  <Step title="Start using it">
    Your new allowance applies as soon as the payment goes through. Check it on [**Settings** > **Usage**](/docs/account/usage).
  </Step>
</Steps>

To move to a smaller plan, or between monthly and yearly billing, click **Change plan**. Stripe shows the new price and when it takes effect before you confirm.

## Update your payment method

Click **Edit payment details**, then add a card or change the default one on the Stripe page.

## Download invoices and receipts

Click **View billing history**. Stripe lists every invoice with a link to download it as a PDF.

## Free trials

If your plan started with a free trial, the card shows how many days are left and the date the trial ends. During a trial you get the plan's full allowance, except image & video credits, which are limited to 50.

To start paying early and unlock the full credit allowance, click **End trial and start plan now**. You're charged right away, a new billing period starts today, and you give up the trial days you had left.

## Cancel your plan

<Steps>
  <Step title="Click Cancel plan">
    It's next to the other buttons on the plan card.
  </Step>

  <Step title="Confirm">
    Read the notice and click **Yes, cancel my plan**, then finish on the Stripe page.
  </Step>
</Steps>

Your plan stays active until the end of the period you've paid for, and the card shows the date it ends. After that, your account moves to the Free plan. Changed your mind before then? Click **Don't cancel** on the same card.

<Warning>
  Some older plans are no longer sold. If you're on one and cancel, you can't subscribe to it again later.
</Warning>

## If a payment fails

If your card is declined, we email you a link to pay the invoice. Until it's paid, your account works like the Free plan. Pay the invoice from that email to restore your plan right away, and update your card with **Edit payment details** so the next renewal goes through.

## Add-ons

* **Team seats.** Add people to your plan from [**Settings** > **Team**](https://easy-peasy.ai/settings/team) with **Purchase Seats**. Each seat gets the plan's full allowance.
* **Always-On Hosting.** Keeps a website or app that Marky built for you running around the clock. Turn it on for each app in its deployment settings.
* **Extra credits and words.** See [Buy more credits](/docs/account/buy-credits).

## Questions

<AccordionGroup>
  <Accordion title="Do unused credits carry over to the next month?">
    No. Your plan allowance resets every month. Credits and words you buy separately do carry over until you use them. See [Plans and credits](/docs/account/plans-and-credits#when-your-allowance-resets).
  </Accordion>

  <Accordion title="I pay yearly. Do I get my whole year of credits at once?">
    No. Yearly plans get the same allowance every month, renewing on the same day each month.
  </Accordion>

  <Accordion title="Can I get a refund?">
    Email [support@easy-peasy.ai](mailto:support@easy-peasy.ai) with your account email and we'll look at it.
  </Accordion>
</AccordionGroup>


# Teams
Source: https://easy-peasy.ai/docs/account/team

Add team members to your plan, buy extra seats, see each member's usage, top up their credits, and share your work with the team.

A team shares one subscription. Every member gets their own login, their own history, and the plan's full allowance of words, credits, and characters each month. The account owner manages seats and billing.

Teams plans include 3 or more seats. The Starter and Unlimited plans (except Unlimited 50) can add seats too. Compare plans on the [pricing page](https://easy-peasy.ai/pricing).

## Your team

Open **Settings** > **Team**.

<Frame>
  <img alt="The Team tab: Team Members, 1/1 seats used, Purchase Seats and Invite a member buttons, a Period menu set to Current Period, and a table with Email, Role, Words Used, Image/Video Credits split into 49 on images and 45 on videos, Additional Credits, and Additional Words" />
</Frame>

The table shows each member's usage for the selected **Period**: this billing period or up to 6 months back.

| Column | What it shows |
| - | - |
| **Words Used** | Words used this period |
| **Image/Video Credits** | Credits used, split into images and videos. Click a number to see those items |
| **Additional Credits**, **Additional Words** | Extra credits and words the member has on top of the plan |

Team members see a message to ask the account owner; only the owner manages the team.

## Invite team members

<Steps>
  <Step title="Click Invite a member">
    It's available while you have a free seat. If all seats are used, [add a seat](#add-or-remove-seats) first.
  </Step>

  <Step title="Invite by email or link">
    Type their email addresses, separated by commas, and click **Send**. Or copy the link under **Invite by link** and send it yourself.
  </Step>

  <Step title="They sign up">
    When they sign up through the invitation, they join your team and appear in the table.
  </Step>
</Steps>

To remove someone, click **Remove Member** on their row and confirm. Their seat becomes free again.

## Add or remove seats

* **Purchase Seats** adds seats to your subscription, up to 50 extra. Each seat costs the same as your plan, and you pay the prorated amount for the rest of the current period.
* **Remove Seats** removes empty seats. Remove the member first. The change is immediate, and the unused time is credited on your next invoice.

On plans that can't add seats, such as Free and Unlimited 50, you'll see **Upgrade to Teams plan** instead.

## Top up a member

Click **Buy Additional Credits or Words** on a member's row, choose **Image & Video Credits** or **Words**, and pay. They go to that member only. See [Buy more credits](/docs/account/buy-credits).

## Share with your team

Most things you create can be shared with the team. Turn on **Share with Team Members**, or the team option in the share window, on:

* [transcriptions](/docs/audio/transcripts#rename-share-and-delete) and [speech](/docs/audio/text-to-speech),
* images and Marky chats,
* [agents](/docs/agents/overview), [workflows](/docs/workflows/share), and [custom tools](/docs/tools/custom-tools#share-a-tool).

Shared items appear in your teammates' lists. On the **Images** and **Videos** tabs of History, the owner can also filter by team member.

<Columns>
  <Card title="Manage your subscription" icon="credit-card" href="/docs/account/subscription">
    Change plan or payment details.
  </Card>

  <Card title="Check your usage" icon="chart-simple" href="/docs/account/usage">
    See your own usage this period.
  </Card>
</Columns>


# Check your usage
Source: https://easy-peasy.ai/docs/account/usage

See how much of this period's allowance you've used, when it resets, and how your usage changes day to day.

Open [**Settings** > **Usage**](https://easy-peasy.ai/settings/usage), or choose **Usage** from the account menu under the logo.

<Frame>
  <img alt="The Usage Statistics card with bars for words, image and video generation, text to speech, transcription, AI agents, and tokens, and the date the limits reset" />
</Frame>

Each bar shows how much you've used this period out of your plan's allowance plus any bonus you bought.

| Bar | What it measures | Links |
| - | - | - |
| **Words Usage** | Words used by Marky Agent, tools, and agents | **View history**, and **Buy more words** on larger plans |
| **Image & Video Generation Usage** | Credits used by images, videos, and other visual features | **View images**, **View videos**, **Buy more credits** |
| **Text to Speech Usage** | Characters used by text to speech and the other audio features | **View history** |
| **AI Transcription Usage** | Files transcribed | **View transcriptions** |
| **AI Agents Usage** | Agents you've created, out of your plan's limit | **Manage agents** |
| **Tokens Usage** | A fair-use limit on the total text AI models process for you | |

Under the bars you'll see the date your limits reset and your current plan, with a link to upgrade. See [Plans and credits](/docs/account/plans-and-credits) for how each balance is counted.

## Daily activity

Below the usage card, **Daily Activity** charts your usage day by day. Pick **Last 7 days**, **Last 14 days**, **Last 30 days**, **Last 90 days**, or a custom range to spot what's using the most.

## About the tokens limit

**Tokens Usage** counts everything the AI models read and write for you, including long conversations and files. It isn't something you spend like words or credits; it's a safety limit against runaway automation, and most people never get close.

If your usage goes far past it, we'll email you. Very heavy usage can pause the account until you change your plan or buy more words.

## Team usage

On a team, each member has their own allowance. The owner can see everyone's words and credits for this period or any of the last six on [**Settings** > **Team**](https://easy-peasy.ai/settings/team).


# Actions
Source: https://easy-peasy.ai/docs/agents/actions

Let your agent do things during a chat: show buttons and forms, collect details, book meetings, search the web, call your API, and send emails.

Actions let your agent do more than answer. With a **Custom Button** it can send a visitor straight to your booking page; with **Collect Data** it can take a visitor's details; with **Google Calendar** it can book a meeting. You describe when each action should be used, and the agent decides during the conversation.

<Note>
  Actions are for paid plans. Each plan allows a number of actions per agent, shown at the top right of the page as **used / allowed**. On the Free plan, the page offers a 7-day trial. See [Limits and usage](/docs/agents/limits-and-usage).
</Note>

## Add an action

<Steps>
  <Step title="Pick an action">
    Open your agent, click **Actions**, and click an action under **Add an action**.

    <Frame>
      <img alt="Add an action: cards for Custom Action, Custom Button, Collect Data, Suggested Messages, Custom Form, Fetch URL, Code Interpreter, Web Search, and Email" />
    </Frame>
  </Step>

  <Step title="Name it and say when to use it">
    Enter an **Action Name** and the **Trigger Instructions**: when the agent should use this action. Be specific, for example "When the user wants to book a class, sign up, or try their free first class". Click **Save**.
  </Step>

  <Step title="Configure it">
    Fill in the action's settings below, such as the button's text and link, and click **Save Configuration**.

    <Frame>
      <img alt="Action Configuration for Book a class: Action Name, Trigger Instructions, and the Custom Button section with Button Text Book a class and the URL of the studio's website" />
    </Frame>
  </Step>

  <Step title="Test it">
    Ask something that should trigger it in the [Playground](/docs/agents/playground).

    <Frame>
      <img alt="The chat: the visitor asks to book their first class, and the agent answers that the first class is free and shows a Book a class button" />
    </Frame>
  </Step>
</Steps>

Your actions are listed under **Your Actions**, each with a switch to turn it on or off and a **Configure** button. To remove one, open it and click **Delete Action**.

<Frame>
  <img alt="Your Actions with a Custom Button action named Book a class, switched on, with a Configure button, and 1 / 8 actions used" />
</Frame>

## Available actions

| Action | What it does | Settings |
| - | - | - |
| **Custom Button** | Shows a button that opens a link | **Button Text**, **URL** |
| **Collect Data** | Asks the visitor for details in the conversation, one or two at a time, and saves them when complete | The fields to ask for (each can be **Required**), **Message after saving**, **Also save to Leads**, and where to **Send submissions to**: a **Webhook URL (optional)** and **Email notifications (optional)** |
| **Custom Form** | Shows a form inside the chat | **Form Title**, **Form Description**, **Form Fields** (text, email, number, text area, select, phone, URL), **Submit Button Text** |
| **Suggested Messages** | Shows tappable replies after the agent's answer | **Auto generate suggested messages**, **Disable input field**, **Show after every message** |
| **Google Calendar** | Checks your free times and books meetings in your calendar. One per agent | Connect Google Calendar, default event title and description, meeting length, and time zone. See [Google Calendar](/docs/agents/integrations#google-calendar) |
| **Web Search** | Searches the web for current information. One per agent | **Allowed Domains (Optional)**, **Include Images** |
| **Fetch URL** | Reads a web page or an API during the chat. One per agent | **Allowed Domains (Optional)** |
| **Code Interpreter** | Runs Python for calculations and charts. One per agent | None |
| **Email** | Sends you an email, for example when a visitor asks to be contacted | **To Email Address**, **Subject Template**, **Message Template** |
| **Custom Action** | Calls your own API, for example to look up an order | **Endpoint URL**, **HTTP Method**, **Headers**, **Payload Template** |

### Collect Data or a lead form?

Both capture visitors' details. **Collect Data** asks naturally in the conversation, works on messaging channels too, and can send what it collects to email or a webhook. The [lead form](/docs/agents/leads) is a fixed form in the website chat that appears after the first answer. Turn on **Also save to Leads** to see Collect Data submissions on the same **Leads** page.

### Email templates

In the **Email** action's templates, `{{user}}` is replaced with the visitor's name and `{{message}}` with the message the agent writes. For example: subject `New message from {{user}}`, message `Message from {{user}}: {{message}}`.

### Custom Action

A **Custom Action** sends a request to your API and gives the agent the response.

* To pass values from the conversation, put `${name}` placeholders in the **Endpoint URL**, for example `https://api.example.com/orders/${order_id}`. The agent fills them in, asking the visitor if it needs to.
* For **POST** and **PUT**, the agent writes the JSON body itself. Describe the body you expect in the **Trigger Instructions**, for example "Send `{"email": "...", "question": "..."}`". The **Payload Template** is sent as written when the agent doesn't supply a body; placeholders in it aren't filled in.
* Add authentication, such as an API key, under **Headers**. The agent never sees your headers.

## How the agent decides

Each action you turn on becomes something the agent can use, with your **Trigger Instructions** as its description. The agent picks an action when the conversation matches them, and can use several in one reply. Clear, specific trigger instructions work best. You can also mention actions in the agent's [instructions](/docs/agents/instructions-and-model#instructions), for example "Always show the Book a class button after explaining prices."

Actions work in the website chat and on messaging channels such as WhatsApp and Telegram. Buttons, forms, and suggested messages are made for the website chat. Actions don't run through the [REST API](/docs/agents/integrations#rest-api).

<Columns>
  <Card title="Leads" icon="address-card" href="/docs/agents/leads">
    Collect and export visitors' details.
  </Card>

  <Card title="Integrations" icon="plug" href="/docs/agents/integrations">
    Google Calendar, Zapier, and the REST API.
  </Card>
</Columns>


# Analytics
Source: https://easy-peasy.ai/docs/agents/analytics

Track your agent's chats, messages, visitor feedback, and conversation sentiment over time.

**Analytics** shows how much your agent is used and how visitors feel about its answers. Open your agent and click **Analytics**.

<Frame>
  <img alt="Agent Analytics for the last 14 days: Total Chats 2, Total Messages 6, Positive Feedback 0, Negative Feedback 0, and a Daily Activity chart of messages and chats" />
</Frame>

Choose a period at the top: the last 7, 14, 30, or 90 days, or **Custom** for any range up to 120 days.

| Section | What it shows |
| - | - |
| **Total Chats**, **Total Messages** | How many conversations and messages there were |
| **Positive Feedback**, **Negative Feedback** | Thumbs up and down visitors gave the agent's answers. Visitors can rate answers when **Collect User Feedback** is on in [Appearance](/docs/agents/appearance#chat-features) |
| **Daily Activity** | Messages, chats, likes, and dislikes per day |
| **Sentiment Distribution** | How many conversations were positive, negative, or neutral |
| **Sentiment Trend Over Time** | How sentiment changed from day to day |

## Use it to improve your agent

* **Negative feedback or negative sentiment** points to answers worth checking. Open the conversation in [**Conversations**](/docs/agents/conversations), and click **Revise answer** to fix it.
* **A jump in messages** after a change, such as a new page on your website, shows whether people find the chat.

<Columns>
  <Card title="Conversations" icon="comments" href="/docs/agents/conversations">
    Read the chats behind the numbers.
  </Card>

  <Card title="Leads" icon="address-card" href="/docs/agents/leads">
    Get a daily summary by email.
  </Card>
</Columns>


# Appearance
Source: https://easy-peasy.ai/docs/agents/appearance

Set your agent's name, greeting, suggested messages, colors, font, chat features, and chat bubble, with a live preview.

**Appearance** controls what visitors see in the chat: the header, the first message, the buttons under the text box, and the colors. The preview on the right updates as you edit. Click **Save** at the bottom to publish your changes to every website chat.

<Frame>
  <img alt="The Appearance page with Agent Avatar, Agent Name, Description, Greeting Message, and Suggested Messages, next to a preview of the chat with a purple header" />
</Frame>

The chat bubble on your website is set here too, under [**Chat bubble**](#chat-bubble): where it sits, its size, icon, color, and text. To put it on your site, copy the install code from [**Channels** > **Website**](/docs/agents/website#add-the-chat-bubble).

## Content

| Setting | What it does |
| - | - |
| **Agent Avatar** | The picture in the chat header, and next to the agent's messages if you turn that on |
| **Agent Name** | The name in the chat header |
| **Description** | The line under the name in the header |
| **Greeting Message** | The first message visitors see |
| **Suggested Messages** | Questions visitors can tap to start, one per line |
| **Suggested messages layout** | **Scroll in one row** keeps the greeting in view; **Show all** wraps every suggestion onto rows |
| **Auto show initial messages pop-ups after (in seconds)** | How long before the greeting pops up next to the closed chat bubble. Enter a negative number to turn this off |
| **Message Placeholder** | The hint text in the message box |

## Chat features

| Switch | What it does |
| - | - |
| **Show Copy Button** | A copy button under the agent's answers |
| **Collect User Feedback** | Thumbs up and down under answers. The counts appear in [Analytics](/docs/agents/analytics) |
| **Allow Message Regeneration** | Lets visitors ask for a new version of an answer |
| **Enable Voice Input** | A microphone button that turns speech into typed text |
| **Allow image attachments** | Lets visitors send screenshots and photos |

### Image attachments

With **Allow image attachments** on, visitors can attach up to 4 images per message (PNG, JPEG, or WebP, up to 5 MB each) in the website chat, on the shareable link, on [Telegram](/docs/agents/telegram), and on [Messenger](/docs/agents/messenger). The agent looks at them and answers.

* The agent's model must be able to see images: GPT, Claude, or Gemini models. With another model, the attach button doesn't appear.
* Each image the agent looks at uses 200 words, on every model. The agent keeps up to 3 recent images in view, so an image is counted again for each reply it's still in view for.

## Style

<Frame>
  <img alt="The Style section: Theme Light, Agent bubble style with background #6366f1 and white text, Header style, Chat font Default, Profile icon size Medium, Show avatar next to agent messages, and Human bubble style" />
</Frame>

| Setting | What it does |
| - | - |
| **Theme** | Light or dark chat |
| **Agent bubble style** | Background and text color of the agent's messages |
| **Header style** | Background color of the header. Empty uses the agent bubble color |
| **Chat font** | The font of the whole chat: a preset such as Inter, or **Custom web font…** with a font name and a stylesheet URL |
| **Profile icon size** | Size of the avatar in the header, from **Small** to **3X Large** |
| **Show avatar next to agent messages** | Shows an avatar beside each answer. **Message Avatar** sets a different picture for it |
| **Human bubble style** | Background and text color of the visitor's messages |

## Chat bubble

The chat bubble is the button that opens the chat on your website. Saved changes reach your website within about a minute, without changing the install code.

| Setting | What it does |
| - | - |
| **Position** | **Bottom Right**, **Bottom Left**, **Top Right**, **Top Left**, or a tab on the edge of the page: **Side Right**, **Side Left** |
| **Chat bubble size** | **Medium**, **Large**, or **Extra Large** (not for side tabs) |
| **Chat window width** | **Default** (400px), **Wide** (480px), or **Extra wide** (560px) on computers. On phones the chat always opens full screen |
| **Chat bubble icon** | Pick one of the built-in icons, or click **+** to upload your own (PNG, JPG, WebP, or GIF, up to 5 MB). An uploaded image can **Fill the bubble**, which suits photos, or **Fit inside**, which suits logos. Agents made before these icons arrived keep **Classic**, the original icon, until you pick another; newer agents start with **Spark**. Side tabs show text only |
| **Chat bubble color** | The bubble's color. Until you pick one, it matches your agent's header color; the reset button goes back to it |
| **Show text in the chat bubble** | Adds a label next to the icon, such as "Chat with us", up to 40 characters, with the icon on the left or the right. It folds away while the chat is open. A side tab always shows its text: set it in **Text on the side tab** |

While the **Appearance** tab of the [Playground](/docs/agents/playground) is open, the bubble appears under the chat preview; click it to see it with the chat open.

The icon and text turn white or dark on their own, so they stay readable on the color you pick. If you pasted an older install code that sets a position, size, width, color, or icon, it keeps working: whatever you save here takes priority over it.

## The "Powered by" footer

**Show 'Powered by Easy-Peasy.AI' footer** can be turned off on the Unlimited 200 and higher plans and on Teams plans.

<Columns>
  <Card title="Website" icon="window-maximize" href="/docs/agents/website">
    Style the chat bubble and copy the code.
  </Card>

  <Card title="Playground" icon="flask" href="/docs/agents/playground">
    Change the look while you chat.
  </Card>
</Columns>


# Basecamp
Source: https://easy-peasy.ai/docs/agents/basecamp

Add your agent to a Basecamp chat so your team can ask it questions with a command.

Your agent can answer in a Basecamp chat (Campfire). Your team types the chatbot's command followed by a question, and the agent replies in the chat. Basecamp is available on every plan.

## Add your agent to Basecamp

<Steps>
  <Step title="Copy the Command URL">
    Open your agent, click **Channels**, and click **Connect** on **Basecamp Chatbots**. Copy the **Command URL**. Keep it private: anyone with it can send questions to your agent.
  </Step>

  <Step title="Add a chatbot in Basecamp">
    In Basecamp, open the chat you want, click the **⋯** menu at the top right, and choose **Configure chatbots**. Click **Add a new chatbot**, give it a name, such as `helper`, an optional avatar, and paste the **Command URL**. Click **Add this chatbot**.
  </Step>

  <Step title="Ask a question">
    In the chat, type `!helper` followed by your question, using the name you chose. The agent replies in a few seconds.
  </Step>
</Steps>

## Good to know

* The agent only answers when someone uses its command; it doesn't post on its own.
* You can add several agents to one chat, each with its own name, such as `!hr` and `!support`.
* Each agent accepts up to 30 commands every 10 minutes from one place.
* To remove it, delete the chatbot in **Configure chatbots**.

<Columns>
  <Card title="Slack" icon="slack" href="/docs/agents/slack">
    Answer your team in Slack.
  </Card>

  <Card title="Channels" icon="share-nodes" href="/docs/agents/channels">
    All the places your agent can answer.
  </Card>
</Columns>


# Channels
Source: https://easy-peasy.ai/docs/agents/channels

Put your agent on your website, WhatsApp, Messenger, Telegram, Slack, LINE, email, phone, and more, and see what works on each channel.

One agent can answer on many channels at once, with the same knowledge, instructions, and actions everywhere. Open your agent and click **Channels** to connect them.

<Frame>
  <img alt="The Channels page: Your website with Chat bubble, Shareable link, and Iframe embed, and Messaging & voice with Telegram, WhatsApp, LINE, and Phone Calls (Twilio)" />
</Frame>

## Available channels

| Channel | How people reach your agent | Plans |
| - | - | - |
| [Website](/docs/agents/website) | A chat bubble, a shareable link, or a chat inside a page | All |
| [Website platforms](/docs/agents/website-platforms) | WordPress, Wix, Squarespace, Webflow, and other site builders | All |
| [Telegram](/docs/agents/telegram) | Messages to your Telegram bot | All |
| [Facebook Messenger](/docs/agents/messenger) | Messages to your Facebook Page | All |
| [Slack](/docs/agents/slack) | Direct messages and @mentions in your workspace | All |
| [WhatsApp](/docs/agents/whatsapp) | Messages to your WhatsApp Business number | Paid |
| [LINE](/docs/agents/line) | Messages to your LINE Official Account | Paid |
| [Email](/docs/agents/email) | Emails to the agent's own address | Paid |
| [Phone calls](/docs/agents/phone-calls) | Calls to a Twilio phone number | Paid |
| [Shopify](/docs/agents/shopify) | The chat on your store, with product search and order status | Paid |
| [Basecamp](/docs/agents/basecamp) | Commands in a Basecamp chat | All |
| [REST API](/docs/agents/integrations#rest-api) and [Zapier](/docs/agents/integrations#zapier) | Your own apps and automations | All |

On the Free plan, the paid channels show **Pro** and an **Upgrade** button.

## Connect a channel

Click **Connect** on a channel's card. Most channels open a window where you sign in or paste a key; each channel's page has the steps. The book icon on a card opens its setup guide.

<Frame>
  <img alt="Messaging & voice channel cards: Telegram, WhatsApp, LINE, Phone Calls (Twilio), Shopify, Email, Facebook Messenger, Slack, WordPress, and Basecamp Chatbots, each with a Connect button" />
</Frame>

Once connected, a card shows **Connected**. **Connected** at the top of the page filters to your connected channels, and the search box finds a channel by name.

## Pause a channel

Open a connected channel and switch off **Enable integration**. The agent stops answering on that channel until you switch it back on, and the card shows **Paused**. Your other channels keep working.

## What works where

| | Website chat | WhatsApp, Messenger, Slack, LINE, Email, Basecamp | Telegram | Phone calls | REST API |
| - | - | - | - | - | - |
| Knowledge and instructions | Yes | Yes | Yes | Yes | Yes |
| [Actions](/docs/agents/actions) | Yes | Yes | Yes | Yes | No |
| [Image attachments](/docs/agents/appearance#image-attachments) | Yes | No | Yes | — | No |
| [Voice calls](/docs/agents/voice) | Yes | No | No | Yes | No |
| [Lead form](/docs/agents/leads) | Yes | No | No | No | No |

Buttons, forms, and suggested messages are made for the website chat. [Conversations](/docs/agents/conversations) lists chats from the website, WhatsApp, Messenger, LINE, and Telegram.

<Columns>
  <Card title="Website" icon="window-maximize" href="/docs/agents/website">
    Start with the chat bubble on your site.
  </Card>

  <Card title="WhatsApp" icon="whatsapp" href="/docs/agents/whatsapp">
    Connect a WhatsApp Business number in one click.
  </Card>
</Columns>


# Conversations
Source: https://easy-peasy.ai/docs/agents/conversations

Read what people ask your agent, see where they came from, fix wrong answers, and export conversations.

**Conversations** shows every chat your agent has had, so you can see what people ask and how the agent answers. It's the fastest way to improve an agent: find a weak answer, correct it, and the agent gets it right next time.

<Frame>
  <img alt="Conversations: a list of two website chats, one marked Positive, and the open conversation with two questions and the agent's answers, each with a confidence score and a Revise answer button" />
</Frame>

## Read conversations

Open your agent and click **Activity** > **Conversations**.

* The menu at the top chooses the channel: **Web**, **WhatsApp**, **Messenger**, **LINE**, or **Telegram**. **Web** includes the website chat, the shareable link, the Playground, and voice calls.
* Each conversation in the list shows the visitor's country, device, and browser, the time, the first message, and its sentiment once it's been analyzed. A phone icon marks conversations with a voice call, with the number of calls when there was more than one.
* Click a conversation to read it. Scroll down and click **Load more conversations** for older ones.
* The refresh icon beside **Conversations** loads new messages without reloading the page.

## Filter conversations

On the **Web** channel, click the filter icon beside **Conversations** to narrow the list. The filters apply as you pick them, appear as chips under the channel menu, and can be removed one at a time or with **Clear all**.

| Filter | Shows |
| - | - |
| **Started** | Conversations started in the last 24 hours, 7, 30, or 90 days |
| **Sentiment** | Positive, neutral, or negative conversations |
| **Feedback** | Conversations where the visitor gave an answer a thumbs up or a thumbs down |
| **Conversations with a voice call** | Conversations where the visitor called your agent |
| **Visitors who shared contact details** | Conversations where you know the visitor's name, email, or phone number |

Under each answer, a percentage shows how confident the agent was. Low scores point to questions its knowledge doesn't cover well.

<Note>
  Conversations on Slack, email, phone calls, Basecamp, and the REST API aren't listed here.
</Note>

## Voice calls

When a visitor calls your agent from the chat widget, the call is part of the same conversation as their typed messages. To list only conversations with a call, turn on **Conversations with a voice call** in the filters.

In the conversation, a marker shows where each call starts, with its time and length, and another where it ends and why: the caller left, the call ended after silence, it reached the time limit, or an error ended it. A call where nobody spoke gets a single marker.

Each message of a call carries a tag:

| Tag | Meaning |
| - | - |
| **Spoken** | The visitor said it, or the agent said its reply out loud |
| **Typed during the call** | The visitor typed it into the call instead of speaking |
| **Spoken and typed** | The visitor's question combined speech and typed text |
| **Interrupted** | The visitor spoke before the agent finished this reply |
| **Voice call** | The message was part of a call made before October 6, 2026, before spoken and typed messages were told apart |

**Details** lists every call with its date, length, number of exchanges, and how it ended, plus the total talk time.

## Conversation details

Click **Details** above the conversation to see who the visitor was and where they came from.

<Frame>
  <img alt="The Details tab: Contact Anonymous, Source Playground, Status Ongoing, Sentiment positive, Messages 4, Country Singapore, Language en-US, Timezone Asia/Singapore, Browser macOS - Chrome, an IP address, Created, Last activity, and Conversation ID" />
</Frame>

| Detail | What it shows |
| - | - |
| **Contact** | The visitor's name or email when they filled in a lead form or your website passed it on, or **Anonymous** |
| **Source** | Website, Playground, or voice call: how the conversation started |
| **Status** | **Ongoing** while the chat is active, then ended |
| **Sentiment**, **Messages** | The mood of the conversation and how many messages it has |
| **Country**, **Language**, **Timezone**, **Browser**, **IP address** | Where the visitor was and what they used |
| **Started on**, **Last page**, **Referrer** | The page the chat started on, the page they were last on, and where they came from |
| **Created**, **Last activity**, **Conversation ID** | When it started, when the last message was sent, and its ID |

To name visitors who are signed in to your website, pass their details to the chat. See [Chat widget for developers](/docs/developers/chat-widget#identify-the-visitor).

## Revise an answer

When the agent gets something wrong, click **Revise answer** under the answer. Write the answer it should have given in **Expected Response** and click **Save**.

<Frame>
  <img alt="The Revise Answer window with the User Message What classes do you offer?, the agent's response, and an Expected Response to edit" />
</Frame>

The question and your answer are added to [**Knowledge**](/docs/agents/knowledge) as a Q\&A, and the answer is marked **Revised**. The agent uses your answer from the next question on.

## Export and delete

* To download the conversations, click the **⋯** menu beside **Conversations** and choose **Export as PDF** or **Export as JSON**. The export contains every conversation of the channel, not only the filtered ones. It isn't available for Telegram.
* To delete a conversation, hover over it in the list and click the trash icon. Deleting can't be undone.

<Columns>
  <Card title="Analytics" icon="chart-line" href="/docs/agents/analytics">
    Chats, messages, and feedback over time.
  </Card>

  <Card title="Leads" icon="address-card" href="/docs/agents/leads">
    Details your visitors left.
  </Card>
</Columns>


# Create an agent
Source: https://easy-peasy.ai/docs/agents/create

Build an agent from your website in a few steps: add your content, choose what the agent does, and open it in the Playground.

The **New AI Agent** page builds a working agent from your website in about a minute. It reads your pages, writes the agent's instructions, name, greeting, and starter questions, and matches the chat colors to your brand. You can change all of it afterwards.

## Create your agent

<Steps>
  <Step title="Start a new agent">
    Open **Agents** in the sidebar and click the button at the top right to create an agent.

    <Frame>
      <img alt="The Agents page with the create button at the top right and one agent, Example Yoga Studio, marked Public, with an Edit button" />
    </Frame>
  </Step>

  <Step title="Add your website and other content">
    Under **How should we train your agent?**, enter your website in **Your website**. The agent reads your pages for its knowledge and takes your logo and colors from them.

    To add more, click **Add files, text, Q\&A and more** under **Other sources**:

    <Frame>
      <img alt="The Add a knowledge source window with Files, Text, Q&A, Sitemap, and YouTube" />
    </Frame>

    | Source | What to add |
    | - | - |
    | **Files** | PDF, DOCX, CSV, TXT, or EPUB documents, up to 10 at a time |
    | **Text** | Anything you type or paste |
    | **Q\&A** | A question and the exact answer you want the agent to give |
    | **Sitemap** | Your sitemap's URL, to add the pages it lists |
    | **YouTube** | A video's URL; the agent learns from its transcript. Add a language code, such as `en`, if it isn't English |

    <Frame>
      <img alt="Step one of New AI Agent: the website example-yoga-demo.easy-peasy.site, a Q&A source Is my first class free?, and a preview of the chat" />
    </Frame>

    A website isn't required: you can build an agent from files, text, or Q\&A alone.
  </Step>

  <Step title="Choose what the agent does">
    Under **What should your agent do?**, pick **Customer support**, **Sales agent**, **Shopping assistant**, or **Something else**. Each fills in a short description of the agent's job. Edit it to fit your business.

    <Frame>
      <img alt="What should your agent do? with Customer support selected and its description: answer customer questions clearly and concisely, stay polite, and escalate billing or account issues to a human" />
    </Frame>
  </Step>

  <Step title="Review the pages found (if shown)">
    When your website has several pages, **Here are the links we found** lists them. Uncheck the ones the agent shouldn't learn from, such as a blog archive or legal pages. The agent learns from up to 100 pages. Fewer, better pages give sharper answers.
  </Step>

  <Step title="Choose where people will talk to it (optional)">
    Under **Where will people talk to it?**, pick the channels you plan to use, such as **Website widget** or **WhatsApp**. This only tailors your getting-started checklist. You connect channels later.

    <Frame>
      <img alt="Where will people talk to it? with channel chips: Website widget (selected), Shareable link, Embed in a page, Slack, Messenger, Telegram, WhatsApp, Email, Shopify, Phone calls, REST API" />
    </Frame>
  </Step>

  <Step title="Click Create my agent">
    The agent is created, reads your website, and matches your branding. Then the [Playground](/docs/agents/playground) opens so you can talk to it.

    <Frame>
      <img alt="We are building a customer support agent for example-yoga-demo.easy-peasy.site, with the steps Creating your agent, Reading your website and matching your branding, and Opening your playground" />
    </Frame>
  </Step>
</Steps>

## What you get

* A name, a one-line description, a greeting, and three suggested questions, written from your website.
* Instructions for the agent: a short description of your business, what you wrote for the job (word for word), and rules for how to answer. Prices and policies aren't copied into them, so the agent answers those from your pages and stays up to date when they change. Edit the instructions on [**Agent Details**](/docs/agents/instructions-and-model).
* Chat colors and an avatar taken from your website, where it has them.
* Your pages and sources in [**Knowledge**](/docs/agents/knowledge). Large websites take a few minutes to finish reading; the agent can answer from what's ready.
* The model **GPT-6 Luna**. You can [change it](/docs/agents/instructions-and-model#models) at any time.

New agents are public, so the chat bubble and the shareable link work as soon as you add them.

## How many agents you can have

Your plan sets how many agents you can have; the Free plan has one. Agents your teammates share with you count too. When you reach the limit, the create button changes to **Upgrade plan**. See [Limits and usage](/docs/agents/limits-and-usage).

<AccordionGroup>
  <Accordion title="My website wasn't read, or only one page was">
    Some websites block automated reading, and single-page sites have only one page to read. Add your content another way: upload documents, paste text, or add your sitemap in [**Knowledge**](/docs/agents/knowledge).
  </Accordion>

  <Accordion title="The agent's name or colors aren't right">
    Change the name, greeting, and colors on [**Appearance**](/docs/agents/appearance), and the instructions on [**Agent Details**](/docs/agents/instructions-and-model).
  </Accordion>
</AccordionGroup>

<Columns>
  <Card title="Test it in the Playground" icon="flask" href="/docs/agents/playground">
    Chat with your agent and adjust it.
  </Card>

  <Card title="Knowledge" icon="book" href="/docs/agents/knowledge">
    Add documents, Q\&A, and more pages.
  </Card>
</Columns>


# Email
Source: https://easy-peasy.ai/docs/agents/email

Give your agent its own email address, forward your support mailbox to it, and optionally reply from your own domain.

Your agent can answer email. It gets its own address; anything sent there gets a reply in the same thread, signed the way you choose. Keep your existing support address and forward it to the agent, or share the agent's address directly.

<Note>
  Email is available on paid plans.
</Note>

## Connect email

<Steps>
  <Step title="Open the Email channel">
    Open your agent, click **Channels**, and click **Connect** on **Email**.
  </Step>

  <Step title="Set how replies look">
    | Field | What it does |
    | - | - |
    | **Display name** | The sender name people see, such as "Example Yoga Studio". Up to 100 characters |
    | **Reply from your own address** | Optional. See [Reply from your own domain](#reply-from-your-own-domain) |
    | **Email signature** | Added to the end of every reply |
    | **AI disclaimer** | A short note under the signature, such as "This reply was written by an AI assistant." |

    Click **Save**.
  </Step>

  <Step title="Copy the agent's address">
    Open the Email window again. **Your agent email address** now shows the agent's address, ending in `@agents.easy-peasy.ai`. Copy it with the button next to it. The address doesn't change.

    <Frame>
      <img alt="The Email Integration window with the agent's email address on agents.easy-peasy.ai, Display name Example Yoga Studio, Reply from your own address, Email signature, AI disclaimer, and the Enable integration switch" />
    </Frame>
  </Step>

  <Step title="Forward your support mailbox (optional)">
    In your email provider, forward your support address to the agent's address:

    * **Gmail and Google Workspace:** **Settings** > **Forwarding and POP/IMAP** > **Add a forwarding address**, or a routing rule in the Admin console.
    * **Outlook and Microsoft 365:** **Settings** > **Mail** > **Forwarding**, or a mail flow rule in the admin center.
    * **Other providers:** look for forwarding or redirect rules.
  </Step>

  <Step title="Send a test email">
    Email a question to the agent's address from a personal address. The reply usually arrives within a minute, in the same thread. Reply again to see the agent keep the context.
  </Step>
</Steps>

## Reply from your own domain

By default, replies come from Easy-Peasy.AI's mail servers with your display name, and replies to them go to the agent. To send from your own address, such as `support@example.com`:

1. Enter it in **Reply from your own address** and click **Save**. Addresses on free mail services such as Gmail can't be used.
2. **Domain verification** shows a TXT record and a CNAME record. Add both at the company that manages your domain's DNS.
3. Click **Verify domain**. DNS changes can take up to 48 hours. When the status shows **Verified**, replies come from your address.

Until then, replies keep coming from the default sender.

## What the agent answers

* Each sender's thread keeps its history, so follow-up emails are answered in context.
* Auto-replies, out-of-office messages, mailing lists, bounces, and spam are ignored, so the agent never gets stuck replying to machines.
* Your [actions](/docs/agents/actions) work in email.

To stop replies, switch off **Enable integration**. Incoming mail is ignored until you switch it back on, and the address stays yours.

<Columns>
  <Card title="Instructions and model" icon="sliders" href="/docs/agents/instructions-and-model">
    Tell the agent what to do when it doesn't know.
  </Card>

  <Card title="Channels" icon="share-nodes" href="/docs/agents/channels">
    All the places your agent can answer.
  </Card>
</Columns>


# Instructions and model
Source: https://easy-peasy.ai/docs/agents/instructions-and-model

Name your agent, write the instructions it follows, and choose the AI model that writes its answers.

**Agent Details** holds the basics of your agent: its name, its avatar, the instructions it follows, and the AI model that answers. Changes apply to every channel as soon as you click **Save**.

<Frame>
  <img alt="Agent Information: Agent Name Example Yoga Studio, the avatar with a Change button, Instructions with a Business Context and How to respond sections (593 of 40,000 characters), and Model GPT-6 Luna with Change model" />
</Frame>

## Name and avatar

* **Agent Name** is shown in the chat header and in your agent list.
* **Avatar** is the picture in the chat header. Click **Change** to upload an image of up to 5 MB.

You can also change both on [**Appearance**](/docs/agents/appearance), along with the greeting and colors.

## Instructions

Instructions tell the agent who it is, who it talks to, and how to answer. The agent follows them in every conversation, on every channel. New agents get instructions written from your website and the job you chose, split into **Business Context** and **How to respond**.

Good instructions cover:

* **Who the agent is.** "You are the front-desk assistant for Example Yoga Studio."
* **What it helps with, and what it doesn't.** "Answer questions about classes, schedules, prices, and bookings. Don't give medical advice."
* **Tone and length.** "Friendly and brief: two or three sentences unless the visitor asks for detail."
* **What to do when it doesn't know.** "If the answer isn't in your knowledge, say so and suggest emailing the studio."
* **When to use an action.** "When someone wants to book, show the Book a class button."

Instructions can be up to 5,000 characters on the Free plan and 40,000 on paid plans. Try changes in the [Playground](/docs/agents/playground) first: you can test new instructions there and save them when you're happy.

<Tip>
  Put facts in [**Knowledge**](/docs/agents/knowledge), not in the instructions. Instructions are for behavior; knowledge is for what the agent knows. A price list in the instructions is easy to forget when prices change.
</Tip>

## Models

The model is the AI that writes your agent's answers. New agents use **GPT-6 Luna**. To switch, click **Change model**, pick a model, and click **Select**, then **Save**.

<Frame>
  <img alt="The Agent Model window with model cards such as GPT-6 Astra, GPT-6 Sol, GPT-6 Luna (selected), GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna, GPT-4o mini, GPT-5.4 Instant, and GPT-5.4 Thinking" />
</Frame>

What to consider:

* **Words.** On paid plans, replies from models marked **Uses words** below count against your plan's words. Models marked **No** answer without using words.
* **Free plan.** Free accounts can use the models marked **Yes** in the Free plan column. The others are dimmed in the picker, with a link to start a trial or upgrade. An agent saved with a paid model on the Free plan runs on GPT-4o mini.
* **Speed and quality.** Larger models answer harder questions better but take longer. For most customer support, a fast model is enough.

### OpenAI

| Model | Free plan | Uses words |
| - | - | - |
| GPT-6 Astra | No | Yes |
| GPT-6 Sol | No | Yes |
| GPT-6 Luna (default) | Yes | Yes |
| GPT-5.6 Sol | No | Yes |
| GPT-5.6 Terra | No | Yes |
| GPT-5.6 Luna | Yes | Yes |
| GPT-4o mini | Yes | No |
| GPT-5.4 Instant | Yes | Yes |
| GPT-5.4 Thinking | No | Yes |
| GPT-5.4 Pro | No | Yes |
| GPT-5 | No | Yes |
| GPT-5 mini | No | Yes |
| GPT-5 nano | No | No |
| GPT-4.1 | No | Yes |
| GPT-4 Turbo | No | Yes |
| o3-mini | No | Yes |
| o3 | No | Yes |
| o4-mini | No | Yes |

### Anthropic

| Model | Free plan | Uses words |
| - | - | - |
| Claude Sonnet 4.5 | No | Yes |
| Claude Sonnet 5 | No | Yes |
| Claude Fable 5.1 | No | Yes |
| Claude Fable 5 | No | Yes |
| Claude Sonnet 4.6 | No | Yes |
| Claude Opus 4.5 | No | Yes |
| Claude Opus 4.6 | No | Yes |
| Claude Opus 5.5 | No | Yes |
| Claude Opus 5 | No | Yes |
| Claude Opus 4.8 | No | Yes |
| Claude Opus 4.7 | No | Yes |

### Google

| Model | Free plan | Uses words |
| - | - | - |
| Gemini Flash 2.0 | Yes | Yes |
| Gemini Flash 2.5 | Yes | Yes |
| Gemini 3 Flash | Yes | Yes |
| Gemini 3.5 Flash | Yes | Yes |
| Gemini 3.6 Flash | Yes | Yes |
| Gemini 3.7 Flash | Yes | Yes |
| Gemini 3.8 Flash | Yes | Yes |
| Gemini 2.5 Pro | No | Yes |
| Gemini 3.1 Pro | No | Yes |

### xAI

| Model | Free plan | Uses words |
| - | - | - |
| Grok 4 | No | Yes |
| Grok 4.3 | No | Yes |
| Grok 4.6 | No | Yes |

### Meta

| Model | Free plan | Uses words |
| - | - | - |
| Muse Spark 1.3 | No | Yes |

### Deepseek

| Model | Free plan | Uses words |
| - | - | - |
| Deepseek V4 Pro | Yes | Yes |
| Deepseek V4 Flash | Yes | No |

### Z.ai

| Model | Free plan | Uses words |
| - | - | - |
| GLM-5 | No | No |
| GLM 5.3 Flash | Yes | No |
| GLM 5.3 | Yes | No |
| GLM 5.2 | Yes | No |

### MiniMax

| Model | Free plan | Uses words |
| - | - | - |
| MiniMax M3 | Yes | No |

### Moonshot

| Model | Free plan | Uses words |
| - | - | - |
| Kimi K2.6 | Yes | Yes |
| Kimi K2.7 Code | Yes | Yes |

### Azure

| Model | Free plan | Uses words |
| - | - | - |
| GPT-4o | No | Yes |

## Temperature

**Temperature** runs from 0 to 1. Lower values (**Reserved**) give consistent, predictable answers, which suits support. Higher values (**Creative**) vary the wording more. New agents start at 0.7. Some newer models, including the GPT-6 family, ignore this setting.

## Share your agent

The **Share** menu at the top gives you quick links:

| Option | What it does |
| - | - |
| **Chat with Marky** | Opens [Marky](/docs/marky/overview) with this agent selected, so you can chat with it there |
| **Via URL** | Opens the agent's shareable chat page |
| **Widget**, **Iframe** | Open the embed codes on the [Website](/docs/agents/website) channel |

## If your agent is paused

A red **Agent Temporarily Blocked** banner means the agent has stopped replying because your plan's allowance ran out. See [Limits and usage](/docs/agents/limits-and-usage#when-an-allowance-runs-out).

<Columns>
  <Card title="Playground" icon="flask" href="/docs/agents/playground">
    Try instructions and models before saving.
  </Card>

  <Card title="Limits and usage" icon="gauge" href="/docs/agents/limits-and-usage">
    What agents use from your plan.
  </Card>
</Columns>


# Integrations
Source: https://easy-peasy.ai/docs/agents/integrations

Let your agent book meetings in Google Calendar, connect it to thousands of apps with Zapier, and talk to it from your own code with the REST API.

Integrations give your agent access to the tools you already use. Open your agent and click **Integrations**.

<Frame>
  <img alt="Integrations: Apps with Google Calendar and Zapier, and Build your own with Custom actions and REST API" />
</Frame>

## Google Calendar

With Google Calendar connected, your agent can offer free times from your calendar and book meetings during a chat. The visitor is added to the event with the email they give, so they get an invite.

<Steps>
  <Step title="Connect your calendar">
    Click **Connect** on **Google Calendar**, then **Connect with Google**. Sign in with the Google account whose calendar should get the bookings and allow access. The page shows the **Connected Account**.
  </Step>

  <Step title="Add the Google Calendar action">
    On [**Actions**](/docs/agents/actions), add **Google Calendar**. In **Trigger Instructions**, say when to offer a meeting, for example "When someone asks for a demo or a consultation. Business hours are 9:00 to 17:00 on weekdays."
  </Step>

  <Step title="Set the defaults">
    Set the default event title and description, the **Default Meeting Duration (minutes)**, and the **Default Timezone** your free times are offered in. Save the configuration.
  </Step>

  <Step title="Book a test meeting">
    In the [Playground](/docs/agents/playground), write "I'd like to book a call next week." The agent offers free times, asks for your name and email, and books the meeting.
  </Step>
</Steps>

The agent offers only times that are free in your calendar, so block time in Google Calendar that shouldn't be booked. To stop bookings, turn off the action; to disconnect the calendar, click **Disconnect** on the Google Calendar page.

Google Calendar is available on paid plans, because it works through an action.

## Zapier

The Easy-Peasy.AI app on [Zapier](https://zapier.com/apps/easy-peasyai/integrations) lets your agent answer inside automated workflows: reply to new emails, answer form submissions, or post answers in a spreadsheet.

<Steps>
  <Step title="Create a Zap with a trigger">
    In Zapier, click **Create Zap** and choose what starts it, such as a new email in Gmail or a new form submission.
  </Step>

  <Step title="Add the Easy-Peasy.AI action">
    Add an action step, search for **Easy-Peasy.AI**, and choose the action that sends a message to your agent. Connect your account with an API key from [**Settings** > **API/MCP**](https://easy-peasy.ai/settings/api).
  </Step>

  <Step title="Map the message and your agent">
    Paste **Your Agent ID** (from [**Channels** > **Website**](/docs/agents/website#your-agent-id)) and map the message to the text from your trigger.
  </Step>

  <Step title="Send the answer on">
    Add a step that uses the agent's answer, such as a Gmail reply or a new row in Google Sheets. Test the Zap and turn it on.
  </Step>
</Steps>

Each Zap run is a separate question: the agent doesn't remember earlier runs. For ongoing conversations, use a channel such as [WhatsApp](/docs/agents/whatsapp) or [Email](/docs/agents/email). Answers use your plan's words like any chat, and Zapier counts its own tasks.

## REST API

Talk to your agent from your own app or backend. Click **Manage** on **REST API** to see your agent's endpoint, try a request with **Test API**, and copy example code in curl, Node, Python, or Go.

<Frame>
  <img alt="The API page with the agent's endpoint URL, the API key, the message, history, stream, and include_sources parameters, a Test API button, and a curl example" />
</Frame>

```bash theme={null}
curl -X POST https://bots.easy-peasy.ai/bot/YOUR_AGENT_ID/api \
  -H "Content-Type: application/json" \
  -H "x-api-key: YOUR_API_KEY" \
  -d '{
    "message": "What classes do you offer?",
    "history": [],
    "stream": false,
    "include_sources": false
  }'
```

| Field | What it does |
| - | - |
| `message` | The question. Required |
| `history` | Earlier messages, each with `role` (`human` or `ai`) and `text`, so the agent answers in context |
| `stream` | `true` streams the answer as it's written |
| `include_sources` | `true` returns the knowledge the answer came from |

The key is your account's API key, from [**Settings** > **API/MCP**](https://easy-peasy.ai/settings/api). Keep it on your server, never in a web page. The API answers from the agent's knowledge and instructions; [actions](/docs/agents/actions) don't run through the API. See the [API reference](/docs/api-reference/endpoint/chat-with-bot) for the full request and response.

## Custom actions

To let your agent call your own API during a chat, for example to look up an order, add a [**Custom Action**](/docs/agents/actions#custom-action).

<Columns>
  <Card title="Actions" icon="bolt" href="/docs/agents/actions">
    Everything your agent can do in a chat.
  </Card>

  <Card title="API reference" icon="square-terminal" href="/docs/api-reference/endpoint/chat-with-bot">
    The agent endpoint in detail.
  </Card>
</Columns>


# Knowledge
Source: https://easy-peasy.ai/docs/agents/knowledge

Add the websites, documents, text, Q&A, sitemaps, and YouTube videos your agent answers from, and keep them up to date.

**Knowledge** is everything your agent knows about your business. When someone asks a question, the agent looks up the most relevant parts of its knowledge and answers from them. The better and more current this content is, the better the answers.

<Frame>
  <img alt="The Knowledge page with Total characters 4,838 of 10,000,000, Last trained 8 minutes ago, Auto-retrain Off, Retrain All, Add new source, and a table of three finished sources: a Q&A, the website's sitemap, and its home page" />
</Frame>

## Add a source

<Steps>
  <Step title="Click Add new source">
    Open your agent, click **Knowledge**, and click **Add new source**.

    <Frame>
      <img alt="Select a data source: Files, Website Crawler, Text, Q&A, Sitemap XML, and Youtube" />
    </Frame>
  </Step>

  <Step title="Choose the type and fill it in">
    | Source | What to enter |
    | - | - |
    | **Files** | PDF, DOCX, CSV, TXT, or EPUB files, up to 10 at a time |
    | **Website Crawler** | A website URL, how deep to follow links, and how many pages to read (up to 100) |
    | **Text** | Any text you type or paste |
    | **Q\&A** | A question and the answer you want the agent to give |
    | **Sitemap XML** | Your sitemap's URL. Up to 100 new pages are added each time |
    | **Youtube** | A video's URL and its language code, such as `en` or `fr`. The agent learns from the transcript |
  </Step>

  <Step title="Click Add">
    The source appears in the table as **PENDING**, then **PROCESSING**, and **FINISHED** when the agent can use it. The table refreshes on its own.
  </Step>
</Steps>

<Tip>
  **Website Crawler** reads 5 pages unless you change **How many links to crawl**. Raise it to read a whole site, or add your **Sitemap XML** instead, which lists your pages directly.
</Tip>

## Sources and their status

Each row shows the source's type, its content or URL, how many characters it adds, and its status:

| Status | Meaning |
| - | - |
| **PENDING** | Waiting to be read |
| **PROCESSING** | Being read |
| **FINISHED** | Ready: the agent answers from it |
| **FAILED** | Couldn't be read, for example a page that blocks automated reading or a video with no transcript |

When a website crawl or a sitemap finishes, each page it found appears as its own row.

* The refresh icon reads a source again. Use it after a page changes, or to retry a failed source.
* The trash icon deletes a source and everything the agent learned from it.
* **Retry Pending** appears when sources are stuck in **PENDING**, and starts them again.

## Keep your knowledge current

Your agent doesn't notice when your website changes. To update it:

* **Retrain All** reads all your website pages again. Files, text, Q\&A, and videos stay as they are; to replace a document, delete it and upload the new version.
* **Auto-retrain** checks your website pages every week and relearns only the pages whose content changed. It's available on Unlimited 500 and higher plans.

**Last trained** shows when the agent last learned from its sources.

## Fix a wrong answer

When the agent gets something wrong in a real conversation, open it in [**Conversations**](/docs/agents/conversations#revise-an-answer) and click **Revise answer**. Your corrected answer is added to **Knowledge** as a Q\&A, so the agent gives it next time. You can also add a **Q\&A** here directly.

## Limits

**Total characters** shows how much of your plan's knowledge allowance this agent uses. Each agent has its own allowance, from 1 million characters on the Free plan up to 250 million. On the Free plan, an agent can also have up to 250 sources. See [Limits and usage](/docs/agents/limits-and-usage).

## How the agent uses its knowledge

Your content is split into short passages. For each message, the agent searches for the passages closest to the question and the conversation so far, and answers from them. That means:

* **Clear, specific pages work best.** A page that answers "How much is a class pass?" in one sentence beats a long page that mentions prices in passing.
* **Q\&A pairs give exact answers.** Use them for the questions you get most, and for anything the agent keeps getting wrong.
* **Conflicting content causes conflicting answers.** Remove outdated pages and documents.

By default, the agent answers only from its knowledge and says when it doesn't know. You can change this in [**Settings**](/docs/agents/settings#answer-only-from-knowledge).

<AccordionGroup>
  <Accordion title="A source says FAILED">
    Click the refresh icon to try again. If a web page keeps failing, the site may block automated reading: copy the page's text into a **Text** source, or upload it as a PDF. A YouTube video fails when it has no transcript in the language code you entered.
  </Accordion>

  <Accordion title="The agent doesn't know something that's in a document">
    Make sure the source is **FINISHED**. Scanned PDFs without selectable text can't be read; upload a text version. Then try asking the same question with different words, and add a **Q\&A** for the answer you expect.
  </Accordion>

  <Accordion title="The crawler read pages I don't want">
    Delete those rows. Next time, add a **Sitemap XML** or pick pages in the website step of a new agent, which lets you uncheck pages.
  </Accordion>
</AccordionGroup>

<Columns>
  <Card title="Test in the Playground" icon="flask" href="/docs/agents/playground">
    Ask questions and see which sources were used.
  </Card>

  <Card title="Instructions and model" icon="sliders" href="/docs/agents/instructions-and-model">
    Tell the agent how to answer.
  </Card>
</Columns>


# Leads and notifications
Source: https://easy-peasy.ai/docs/agents/leads

Collect visitors' names, emails, and phone numbers in the chat, export them, and get an email for every new lead and a daily summary.

Your agent can collect contact details from visitors, so you can follow up. There are two ways to do it:

* **The lead form**, a short form in the website chat that appears after the agent's first answer. Set it up on the **Leads** page.
* **The [Collect Data](/docs/agents/actions#available-actions) action**, which asks for details during the conversation and works on messaging channels too.

<Note>
  The lead form and notifications are for paid plans.
</Note>

## Set up the lead form

<Steps>
  <Step title="Open Leads">
    Open your agent and click **Leads** in its menu.

    <Frame>
      <img alt="Agent Leads: the Display Leads Form switch, and the fields Title (Let us know how to contact you), Name Field Title (Your Name), Phone Field Title (Your Phone), Email Field Title (Your Email), and Submit Button Text (Submit)" />
    </Frame>
  </Step>

  <Step title="Turn on Display Leads Form">
    Change the **Title**, the field labels, and the **Submit Button Text** if you like. Leave a field's label empty to hide that field: for example, clear **Phone Field Title** to ask only for a name and email.
  </Step>

  <Step title="Click Save">
    The form appears in the chat after the agent's first answer. Visitors can close it, but it comes back after the next answer until they submit it. Submitted details are also sent to the agent as a message.
  </Step>
</Steps>

## See collected leads

Open **Activity** > **Leads**, or click **View collected leads** on the **Leads** page. Each lead shows the date and the details the visitor left, 25 per page. Details collected by the **Collect Data** action appear here too when its **Also save to Leads** option is on.

Click **Export CSV** to download them for your CRM or mailing tool.

## Get notified

Open **Notifications** to choose which emails you get:

<Frame>
  <img alt="Agent Notifications: New Leads Notifications and Daily Summary switches, Notification Email, Lead Email Subject, and Lead Email Template" />
</Frame>

| Setting | What it does |
| - | - |
| **New Leads Notifications** | An email each time a visitor submits the lead form |
| **Daily Summary** | An email each morning with the previous day's conversations and messages, and a link to them. Nothing is sent on days without messages |
| **Notification Email** | Where to send the emails. Separate several addresses with commas. Empty uses your account email |
| **Lead Email Subject** | The subject of lead emails. `{{bot_name}}` is replaced with the agent's name |
| **Lead Email Template** | The body of lead emails, in HTML. Use `{{bot_name}}`, `{{lead_name}}`, `{{lead_email}}`, and `{{lead_phone}}` |

The daily summary covers one UTC day and arrives around 07:00 UTC.

<Columns>
  <Card title="Actions" icon="bolt" href="/docs/agents/actions">
    Collect details in the conversation with Collect Data.
  </Card>

  <Card title="Conversations" icon="comments" href="/docs/agents/conversations">
    See who your leads talked to the agent about.
  </Card>
</Columns>


# Limits and usage
Source: https://easy-peasy.ai/docs/agents/limits-and-usage

How many agents, actions, and knowledge your plan includes, what agent replies use from your allowance, and what happens when it runs out.

## What each plan includes

| Plan | Agents | Actions per agent | Knowledge characters per agent |
| - | - | - | - |
| Free | 1 | 0 | 1M |
| Starter | 2 | 3 | 5M |
| Unlimited 50 | 3 | 5 | 5M |
| Unlimited | 4 | 8 | 10M |
| Unlimited 200 | 8 | 16 | 20M |
| Unlimited 350 | 12 | 32 | 30M |
| Unlimited 500 | 16 | 64 | 40M |
| Unlimited 1000 | 32 | 128 | 50M |
| Unlimited 2000 | 64 | 256 | 100M |
| Unlimited 5000 | 160 | 640 | 250M |
| Teams | 8 | 3 | 8M |

Agents that teammates share with you count toward your number of agents. Compare plans on the [pricing page](https://easy-peasy.ai/pricing).

## What agent replies use

Agents don't use image & video credits. On paid plans:

* **Text replies** use your plan's words when the agent's model is marked **Uses words** in the [model list](/docs/agents/instructions-and-model#models). Models marked **No** don't use words. It's the same allowance as Marky and the tools, plus any [bonus words](/docs/account/buy-credits) you bought.
* **Images** that visitors send use 200 words each time the agent looks at them, on every model. See [Image attachments](/docs/agents/appearance#image-attachments).
* **Voice calls** from the website chat use words at twice the rate of a text reply, on every model. See [Voice calls](/docs/agents/voice#what-calls-cost).
* **Phone calls** through Twilio use words like text replies. Twilio bills the number and minutes separately.

Channels, knowledge, and conversations don't cost anything extra. Check your words on [**Settings** > **Usage**](/docs/account/usage).

## The Free plan

On the Free plan you can build one agent and try it out:

| | Free plan |
| - | - |
| Messages | 100 every 30 days, across all your agents |
| Knowledge | 1 million characters and 250 sources; up to 100 pages per website crawl |
| Instructions | Up to 5,000 characters (40,000 on paid plans) |
| Models | The models marked **Yes** under Free plan in the [model list](/docs/agents/instructions-and-model#models) |
| Channels | Website, shareable link, Telegram, Messenger, Slack, and Basecamp |
| Actions, lead form, notifications | Paid plans. Actions can be tried with a 7-day trial |
| Voice calls | 10 minutes of testing in the Playground every 30 days |
| "Powered by" footer | Always shown |

## When an allowance runs out

Usage is checked once a day.

* **Paid plans:** when your words are used up, agents on models that use words stop replying. Agents on models that don't use words keep working. They start again when your plan renews, when you upgrade, or when you buy more words.
* **Free plan:** after 100 messages in 30 days, all your agents stop replying until you upgrade.

You get an email when an agent is paused, and **Agent Details** shows **Agent Temporarily Blocked**. While paused, visitors get an error instead of an answer.

## If your subscription ends

When you no longer have a paid plan, agents on paid-only models switch to GPT-4o mini, and the "Powered by Easy-Peasy.AI" footer is shown again. You get an email about each change.

<Columns>
  <Card title="Plans and credits" icon="coins" href="/docs/account/plans-and-credits">
    How words and credits work across Easy-Peasy.AI.
  </Card>

  <Card title="Instructions and model" icon="sliders" href="/docs/agents/instructions-and-model#models">
    Which models use words.
  </Card>
</Columns>


# LINE
Source: https://easy-peasy.ai/docs/agents/line

Connect your agent to a LINE Official Account so it answers messages on LINE.

Your agent can answer messages sent to your LINE Official Account. You copy two keys from the LINE Developers console and paste them into Easy-Peasy.AI; the webhook is set up for you.

<Note>
  LINE is available on paid plans. You need a LINE Official Account with the Messaging API turned on.
</Note>

## Connect LINE

<Steps>
  <Step title="Turn on the Messaging API">
    In LINE Official Account Manager, open your account's **Settings** > **Messaging API** and enable it. This creates a Messaging API channel you can open in the [LINE Developers console](https://developers.line.biz/console/).
  </Step>

  <Step title="Copy the channel secret and access token">
    In the LINE Developers console, open your channel:

    * On the **Basic settings** tab, copy the **Channel secret**.
    * On the **Messaging API** tab, under **Channel access token (long-lived)**, click **Issue** and copy the token.
  </Step>

  <Step title="Paste them into Easy-Peasy.AI">
    Open your agent, click **Channels**, and click **Connect** on **LINE**. Paste the **Channel access token** and **Channel secret**, and click **Save**. The keys are checked and the webhook is registered with LINE automatically.
  </Step>

  <Step title="Check the webhook and auto-replies">
    Back on the **Messaging API** tab in the LINE Developers console, make sure **Use webhook** is on. In LINE Official Account Manager, turn off the account's automatic greeting and reply messages if you don't want them to answer alongside your agent.
  </Step>

  <Step title="Send a test message">
    Once connected, the LINE window has a button to open your account in LINE. Add it as a friend and send a question.
  </Step>
</Steps>

## How it works

* The agent answers in one-to-one chats and in groups the account has been added to. Each chat keeps its own history.
* It reads text messages. For other messages, it replies that it can only read text.
* Your [actions](/docs/agents/actions) work on LINE, and conversations appear in [**Conversations**](/docs/agents/conversations) under **LINE**.
* Each LINE Official Account can be connected to one agent.

To stop replies, switch off **Enable integration** in the LINE window.

<Columns>
  <Card title="Channels" icon="share-nodes" href="/docs/agents/channels">
    All the places your agent can answer.
  </Card>

  <Card title="Conversations" icon="comments" href="/docs/agents/conversations">
    Read your LINE chats.
  </Card>
</Columns>


# Facebook Messenger
Source: https://easy-peasy.ai/docs/agents/messenger

Connect your agent to your Facebook Page so it answers Messenger messages automatically.

Your agent can reply to messages sent to your Facebook Page, from Facebook, the Messenger app, or an `m.me` link. You connect it by signing in with Facebook; no developer account or app review is needed. Messenger is available on every plan, including Free.

<Note>
  You need a Facebook Page and admin access to it. Personal profiles can't be connected.
</Note>

## Connect Messenger

<Steps>
  <Step title="Start the connection">
    Open your agent, click **Channels**, and click **Connect** on **Facebook Messenger**. Facebook's sign-in window opens.
  </Step>

  <Step title="Choose your Page">
    Sign in with the Facebook account that manages your Page, approve the permissions, and select the Pages the agent should answer for.
  </Step>

  <Step title="Send a test message">
    Message your Page from another account, or open `m.me/YourPageName`, and ask a question.
  </Step>
</Steps>

Each Page answers through one agent. If you connect a Page that's already on another agent, it's paused there.

## What the agent answers

| Someone sends | The agent |
| - | - |
| A message or a quick reply | Answers it |
| A tap on **Get Started** | Greets them and says what it can help with |
| A photo | Reads it when **Allow image attachments** is on in [Appearance](/docs/agents/appearance#image-attachments) and the agent's model can see images; otherwise asks them to describe it |
| A voice message, video, or file | Asks them to type their question |
| A thumbs-up | Doesn't reply |

While the agent writes, the person sees **Seen** and typing dots. Replies are plain text, because the Messenger apps show formatting symbols like `**` as they are, and long answers arrive as several messages.

Your [actions](/docs/agents/actions) work on Messenger, and conversations appear in [**Conversations**](/docs/agents/conversations) under **Messenger**.

## Take over a conversation

Reply from your Page inbox in Meta Business Suite. The agent steps back in that conversation and stays quiet until an hour has passed since your last reply there; then it answers again. Other conversations aren't affected.

Instant replies and away messages you set up in Meta Business Suite don't count as taking over.

To stop the agent on the whole Page, switch the Page off on the manage page.

## Manage your Pages

Open the Messenger card to see **Manage Facebook Pages**: every connected Page with its status and a switch to pause it. **Connect a New Facebook Page** adds another Page.

If Facebook disconnects a Page, for example after you change your Facebook password or remove Easy-Peasy.AI from your Business Integrations, the Page shows **RECONNECT NEEDED** and we email you. Click **Reconnect** and sign in with Facebook again.

## Tips

* Train the agent on the questions your Page inbox gets most: opening hours, prices, bookings, shipping.
* Tell it in the [instructions](/docs/agents/instructions-and-model#instructions) what to do when it doesn't know, for example "Say a team member will reply during business hours."
* Let people know they're talking to an AI assistant, for example with "Introduce yourself as our AI assistant" in the instructions. Meta requires this in some places, such as California and Germany.

<AccordionGroup>
  <Accordion title="The agent doesn't answer">
    Check that the Page shows **ACTIVE** on the manage page. If it shows **RECONNECT NEEDED**, click **Reconnect**. The agent also stays quiet for an hour in a conversation where someone replied from the Page inbox.
  </Accordion>

  <Accordion title="Two replies arrive for each message">
    Another app or assistant on your Page is answering too, such as a chatbot app or an AI assistant you turned on in Meta Business Suite. Turn the other one off.
  </Accordion>
</AccordionGroup>

Instagram direct messages aren't an agent channel.

<Columns>
  <Card title="WhatsApp" icon="whatsapp" href="/docs/agents/whatsapp">
    Connect a WhatsApp Business number.
  </Card>

  <Card title="Channels" icon="share-nodes" href="/docs/agents/channels">
    All the places your agent can answer.
  </Card>
</Columns>


# Agents overview
Source: https://easy-peasy.ai/docs/agents/overview

Build an AI agent trained on your website and documents, test it, and put it on your website, WhatsApp, Telegram, email, phone, and other channels.

An agent answers your customers' questions from your own content. You give it your website, documents, and FAQs, tell it how to behave, and put it where your customers already are: a chat bubble on your website, a shareable link, WhatsApp, Messenger, Telegram, Slack, LINE, email, or a phone number. Agents were called Bots before.

<Frame>
  <img alt="The Playground of an agent called Example Yoga Studio: the agent's menu on the left, settings with the model GPT-6 Luna, temperature, and instructions in the middle, and a chat on the right where the agent lists the studio's classes and confirms that the first class is free" />
</Frame>

## How it works

<Steps>
  <Step title="Create an agent">
    Enter your website, say what the agent should do, and the agent is built for you: it reads your pages, writes its instructions, and matches your brand colors. See [Create an agent](/docs/agents/create).
  </Step>

  <Step title="Train it on your content">
    Add documents, text, questions and answers, a sitemap, or YouTube videos in [**Knowledge**](/docs/agents/knowledge). The agent answers from what you add here.
  </Step>

  <Step title="Test it in the Playground">
    Chat with your agent, try other models and instructions, and save the setup you like. See [Playground](/docs/agents/playground).
  </Step>

  <Step title="Make it yours">
    Set the [instructions and model](/docs/agents/instructions-and-model), the [look of the chat](/docs/agents/appearance), and [actions](/docs/agents/actions) such as booking a meeting or collecting a lead.
  </Step>

  <Step title="Put it on your channels">
    Add the chat bubble to your [website](/docs/agents/website), or connect [WhatsApp, Telegram, email, and other channels](/docs/agents/channels).
  </Step>

  <Step title="Watch how it does">
    Read [conversations](/docs/agents/conversations), improve answers, and check [analytics](/docs/agents/analytics) and [leads](/docs/agents/leads).
  </Step>
</Steps>

## Find your way around

Open **Agents** in the sidebar to see your agents. Click **Edit** on an agent to open it. Its menu has these pages:

| Page | What it's for |
| - | - |
| **Agent Details** | Name, avatar, [instructions, and model](/docs/agents/instructions-and-model) |
| **Playground** | [Test the agent](/docs/agents/playground) and try other settings |
| **Knowledge** | The [content the agent answers from](/docs/agents/knowledge) |
| **Appearance** | How the [chat looks](/docs/agents/appearance): name, greeting, colors, suggested messages |
| **Voice** | [Voice calls](/docs/agents/voice) with your agent |
| **Activity** | [Conversations](/docs/agents/conversations) and collected [leads](/docs/agents/leads) |
| **Leads** | The [lead form](/docs/agents/leads) shown in the chat |
| **Notifications** | [Emails](/docs/agents/leads#get-notified) about new leads and a daily summary |
| **Channels** | Where [people talk to your agent](/docs/agents/channels) |
| **Integrations** | [Google Calendar, Zapier, and the REST API](/docs/agents/integrations) |
| **Analytics** | [Chats, messages, feedback, and sentiment](/docs/agents/analytics) over time |
| **Actions** | [Things your agent can do](/docs/agents/actions) during a chat |
| **Settings** | [Visibility, allowed domains, sharing, and deleting](/docs/agents/settings) |

The **Getting started** card under the menu lists what's left to set up. You can close it.

## What agents cost

Agents don't use image & video credits. What they use depends on your plan and the agent's model:

* **On paid plans**, replies from most models use your plan's words, the same allowance as Marky and the tools. Some models don't use words at all. See [Instructions and model](/docs/agents/instructions-and-model#models).
* **On the Free plan**, you get one agent that answers up to 100 messages every 30 days.

Your plan also sets how many agents you can have, how many actions each can use, and how much knowledge each can hold. See [Limits and usage](/docs/agents/limits-and-usage).

<Columns>
  <Card title="Create an agent" icon="wand-magic-sparkles" href="/docs/agents/create">
    Build your first agent from your website.
  </Card>

  <Card title="Add the chat to your website" icon="window-maximize" href="/docs/agents/website">
    Copy one line of code, or share a link.
  </Card>
</Columns>


# Phone calls
Source: https://easy-peasy.ai/docs/agents/phone-calls

Give your agent a phone number with Twilio so callers can talk to it.

Your agent can answer a real phone number. Callers hear your greeting, ask their questions out loud, and the agent answers in a natural voice from the same knowledge as in chat. They can interrupt it, just like a person. The number lives in your own Twilio account.

For voice calls from your website chat, see [Voice calls](/docs/agents/voice).

<Note>
  Phone calls are available on paid plans. You also need a Twilio account and a phone number that supports voice. Twilio bills the number and call minutes to your Twilio account.
</Note>

## Connect a phone number

<Steps>
  <Step title="Copy your Twilio keys">
    Sign in to the [Twilio Console](https://console.twilio.com). Under **Account Info** on the home page, copy the **Account SID** (it starts with `AC`) and the **Auth Token**.
  </Step>

  <Step title="Get a voice number">
    In the Twilio Console, go to **Phone Numbers** > **Manage** > **Buy a number**, check **Voice**, and buy a number. You can also use a number you already have in Twilio.
  </Step>

  <Step title="Turn on Twilio's AI features">
    In the Twilio Console, go to **Voice** > **Settings** > **General**. Under **Predictive and Generative AI/ML Features Addendum**, select **Enabled** and click **Save**. Without this, calls end right after they start.
  </Step>

  <Step title="Connect the number">
    In Easy-Peasy.AI, open your agent, click **Channels**, and click **Connect** on **Phone Calls (Twilio)**. Fill in:

    | Field | What to enter |
    | - | - |
    | **Account SID** | From the Twilio Console |
    | **Auth Token** | From the Twilio Console |
    | **Twilio phone number** | In international format, such as `+15551234567` |
    | **Welcome greeting** | What the agent says when it answers, such as "Hi, you've reached Example Yoga Studio. How can I help?" |
    | **Language** | The callers' language as a code, such as `en-US`, `en-AU`, or `de-DE` |

    Click **Save**. Your keys and number are checked, and the number's voice settings in Twilio are updated for you.
  </Step>

  <Step title="Call your agent">
    Dial your Twilio number. You hear your greeting; then ask a question.
  </Step>
</Steps>

On a Twilio trial account, Twilio plays a short message first and asks you to press a key, and only the phone number you verified with Twilio can call. Upgrading your Twilio account removes both limits.

## How it works

* The agent keeps answers short and suited to speech, and remembers what was said earlier in the call.
* Your [actions](/docs/agents/actions) work on calls.
* Each phone number can be connected to one agent.
* Replies use your plan's words like chat replies. Twilio charges for the number and the call minutes in your Twilio account.

To stop answering calls, switch off **Enable integration**. Callers then hear a message that the assistant is paused.

<AccordionGroup>
  <Accordion title="The call ends right after it starts">
    Turn on the **Predictive and Generative AI/ML Features Addendum** in Twilio (step 3). On a trial account, press a key when Twilio's trial message plays.
  </Accordion>

  <Accordion title="The agent mishears callers">
    Set **Language** to your callers' locale, for example `en-GB` or `en-AU`, and click **Save**.
  </Accordion>
</AccordionGroup>

<Columns>
  <Card title="Voice calls" icon="microphone" href="/docs/agents/voice">
    Calls from your website chat.
  </Card>

  <Card title="Channels" icon="share-nodes" href="/docs/agents/channels">
    All the places your agent can answer.
  </Card>
</Columns>


# Playground
Source: https://easy-peasy.ai/docs/agents/playground

Chat with your agent, try other models, instructions, and chat styles side by side, and save the setup you like.

The **Playground** is where you test your agent before your customers see it. The chat on the right is the same chat your website visitors get. The settings on the left let you try another model, other instructions, or a new look, and see the effect right away.

<Frame>
  <img alt="The Playground: Settings with the Chat, Appearance, and Voice tabs, the model GPT-6 Luna, temperature 0.7, and instructions; the chat preview with two questions and the agent's answers; Preview and Deploy buttons at the top right" />
</Frame>

## Test your agent

Open your agent and click **Playground**. Type a question in the chat, or click one of the suggested messages.

* **Show Sources** under the chat lists the knowledge the last answer came from. Use it to check whether an answer came from the right page.
* Under each answer you can copy it, rate it, or regenerate it.
* The refresh icon in the chat header starts a new conversation.

Ask the questions your customers ask most. When an answer is wrong or missing, fix it in [**Knowledge**](/docs/agents/knowledge) or in the instructions.

<Note>
  Playground chats are real conversations: they appear in [**Conversations**](/docs/agents/conversations) and use your plan's allowance like any other chat.
</Note>

## Try a different setup

The **Chat** tab has the settings that change how the agent answers:

| Setting | What it does |
| - | - |
| **Model** | The AI model that writes the answers. See [models](/docs/agents/instructions-and-model#models) |
| **Temperature** | Lower (**Precise**) keeps answers consistent; higher (**Creative**) makes them more varied |
| **Instructions** | How the agent behaves: its role, tone, and rules |
| **Actions** | Turn [actions](/docs/agents/actions) on or off for this test only |

These changes apply only to your Playground chat. When they differ from the agent's saved setup, a bar appears at the bottom with **Unsaved setup**:

* **Save to agent** saves the model, temperature, and instructions to your agent. Every channel uses them from your next message.
* **Discard** goes back to the saved setup. **Reset** at the top does the same and also clears the chat.

The **Actions** switches are never saved. To turn an action off for real, use the [**Actions**](/docs/agents/actions) page.

## Change the look

The **Appearance** tab has the same settings as the [**Appearance**](/docs/agents/appearance) page: name, greeting, suggested messages, colors, and more. The chat preview updates as you type. Click **Save** at the bottom of the tab to keep your changes.

Under [**Chat bubble**](/docs/agents/appearance#chat-bubble), choose where the button that opens the chat on your website sits, and its size, icon, color, and text. While the tab is open, the bubble appears under the chat preview on the side you chose; click it to see how it looks while the chat is open.

<Frame>
  <img alt="The Appearance tab in the Playground with Agent Avatar, Agent Name, Description, Greeting Message, and Suggested Messages, next to the chat preview" />
</Frame>

## Set up voice calls

The **Voice** tab turns on voice calls and sets the voice, the model used on calls, and the call instructions. Changes here save automatically. With voice on, a **Talk** button appears in the chat. See [Voice calls](/docs/agents/voice).

<Frame>
  <img alt="The Voice tab in the Playground with Voice mode and Allow text input during calls switched on and the voice model Same as chat, next to the chat with a call button in the message box" />
</Frame>

## Preview it on a web page

**Preview** opens your agent's real chat bubble on a sample web page, the way it looks on your website. Switch between a light and a dark page to check your colors.

<Frame>
  <img alt="The Preview page: a sample web page with the agent's chat open on the right, showing the greeting and suggested messages" />
</Frame>

## Deploy

**Deploy** lists the fastest ways to put your agent live:

<Frame>
  <img alt="The Deploy menu: Website widget, Website iframe, Shareable link, Shopify, WordPress, and Browse all channels" />
</Frame>

| Option | What it does |
| - | - |
| **Website widget** | A chat bubble on your website. See [Website](/docs/agents/website) |
| **Website iframe** | The chat inside a page instead of floating over it |
| **Shareable link** | A hosted chat page you can link to |
| **Shopify** | The chat bubble plus product and order lookups on your store. See [Shopify](/docs/agents/shopify) |
| **WordPress** | The chat bubble on a WordPress site. See [Website platforms](/docs/agents/website-platforms#wordpress) |
| **Browse all channels** | WhatsApp, Telegram, email, phone, and more. See [Channels](/docs/agents/channels) |

<Columns>
  <Card title="Instructions and model" icon="sliders" href="/docs/agents/instructions-and-model">
    Write better instructions and pick a model.
  </Card>

  <Card title="Channels" icon="share-nodes" href="/docs/agents/channels">
    Put your agent where your customers are.
  </Card>
</Columns>


# Agent settings
Source: https://easy-peasy.ai/docs/agents/settings

Control who can chat with your agent, whether it answers only from its knowledge, which websites can embed it, team sharing, publishing to Marky, and deleting it.

Open your agent and click **Settings**. Click **Save** after changing the settings in the first card.

<Frame>
  <img alt="Agent Settings with Is agent publicly available? on, Should the agent use only knowledge when responding? on, Share with Team Members with an Upgrade to Teams link, and Allowed Domains with example entries" />
</Frame>

## Public access

**Is agent publicly available?** is on for new agents. Leave it on: the chat bubble, the shareable link, and your other website chats need it. With it off, the agent currently answers only in the Playground. **Open in new tab** opens your agent's shareable chat page.

## Answer only from knowledge

**Should the agent use only knowledge when responding?** is on for new agents.

* **On:** the agent searches its [knowledge](/docs/agents/knowledge) for every message and answers from it. When the answer isn't there, it says it doesn't know instead of guessing, and it declines questions unrelated to your content.
* **Off:** the agent doesn't search its knowledge at all. It answers from its instructions and the model's general knowledge, like a general assistant.

Keep it on for customer support. Turn it off only for an agent that shouldn't use documents, such as a general writing helper.

## Share with your team

On a Teams plan, **Share with Team Members** lets everyone on your team view and use the agent. Only the agent's owner can change this. On other plans, **Upgrade to Teams** is shown instead.

## Allowed domains

**Allowed Domains** limits which websites can show your agent's chat. Enter one domain per line, without commas:

```text theme={null}
example.com
*.example.com
localhost:3000
```

`*.example.com` covers every subdomain. Leave the box empty to allow every website.

<Warning>
  With a list set, the agent answers only chats that come from those websites. The [shareable link](/docs/agents/website#shareable-link) stops working, and so does **Preview**, unless you also add `easy-peasy.ai`. Allowed domains keep your widget off other people's websites; they aren't a way to protect private information.
</Warning>

## Publish to Marky Agent

**Publish to Marky Agent** lets other Easy-Peasy.AI users find your agent in [Marky](/docs/marky/overview) and chat with it. Your agent's name, description, and avatar become visible to them, and they use their own allowance for those conversations. Add an **Author name** to show who made it.

<Frame>
  <img alt="Publish to Marky Agent: make this agent available to all users in the Marky Agent chat, with a switch and the status Not published" />
</Frame>

## Delete an agent

Under **Danger Zone**, click **Delete Agent**, type `delete`, and confirm. The agent, its knowledge, and its conversations are deleted and can't be restored. Remove its code from your website and disconnect its channels first, so visitors don't see a broken chat.

<Columns>
  <Card title="Website" icon="window-maximize" href="/docs/agents/website">
    Embed the chat on your site.
  </Card>

  <Card title="Knowledge" icon="book" href="/docs/agents/knowledge">
    What the agent answers from.
  </Card>
</Columns>


# Shopify
Source: https://easy-peasy.ai/docs/agents/shopify

Add your agent to your Shopify store so it answers shoppers, finds products from your catalog, and tells customers where their order is.

Connecting Shopify does two things:

* **Adds the chat to your store**, on every page.
* **Connects your catalog and orders**, so the agent can search your products, with prices and links, and look up an order's status when the customer gives the order number and the email used at checkout.

You create a private app in your own Shopify account; nothing is installed from the Shopify App Store, and you can revoke access at any time.

<Note>
  Shopify is available on paid plans.
</Note>

## Create an app in Shopify

<Steps>
  <Step title="Open the Dev Dashboard">
    In your Shopify admin, go to **Settings** > **Apps and sales channels** > **Develop apps** and click **Build apps in Dev Dashboard**.
  </Step>

  <Step title="Create the app">
    Click **Create app**, choose **Start from Dev Dashboard**, name it, for example "Easy-Peasy.AI agent", and click **Create**.
  </Step>

  <Step title="Choose its permissions">
    Under **API access**, click **Select scopes** and turn on `read_products` and `read_orders`. Add `read_all_orders` to let the agent find orders older than 60 days. Click **Done**, then **Release**.
  </Step>

  <Step title="Install it and copy the credentials">
    On the app's **Overview**, click **Install app**, pick your store, and click **Install**. Then open the app's **Settings** and copy the **Client ID** and **Client secret**.
  </Step>
</Steps>

Older stores that still have **Develop apps** > **Create an app** can create a custom app there instead, with the same permissions, and copy its Admin API access token (it starts with `shpat_`).

## Connect your store

<Steps>
  <Step title="Open the Shopify channel">
    In Easy-Peasy.AI, open your agent, click **Channels**, and click **Connect** on **Shopify**.
  </Step>

  <Step title="Enter your store's details">
    Enter your **Shop domain**, the one ending in `.myshopify.com` (find it in Shopify under **Settings** > **Domains**). Paste the **Client ID** and **Client secret**, or the **Admin API access token (older stores)**. Click **Save**. The connection is checked with Shopify before it's saved.
  </Step>

  <Step title="Add the chat to your theme">
    Copy the **Chat widget embed code** from the same window. In Shopify, go to **Online Store** > **Themes**, click **…** next to your theme, and choose **Edit code**. Open `layout/theme.liquid`, paste the code just before `</body>`, and click **Save**.
  </Step>
</Steps>

To style the bubble (color, position, size), copy the install snippet from [**Channels** > **Website**](/docs/agents/website) instead; it works the same way in `theme.liquid`. If you publish another theme later, paste the code there too.

## Try it

* Ask "Do you have any yoga mats?" The agent lists matching products with prices and links.
* Ask "Where is my order?" The agent asks for the order number and the email used at checkout, and shares the status only when both match.

Product search and order status work in the chat on your website, the shareable link, and the Playground.

## Pause or disconnect

* **Pause:** switch off **Enable integration**. The agent stops using your store data; the chat stays on your theme until you remove the code.
* **Disconnect:** uninstall the app in Shopify, which revokes access immediately, and remove the code from your theme.

<Columns>
  <Card title="Add your agent to your website" icon="window-maximize" href="/docs/agents/website">
    Style the chat bubble.
  </Card>

  <Card title="Knowledge" icon="book" href="/docs/agents/knowledge">
    Add your shipping and returns policies.
  </Card>
</Columns>


# Slack
Source: https://easy-peasy.ai/docs/agents/slack

Add your agent to a Slack workspace so your team can ask it questions in direct messages and channels.

Your agent can answer your team in Slack: in direct messages, and in channels when someone @mentions it. It's a quick way to give everyone access to your handbook, policies, or product docs. Slack is available on every plan, including Free.

<Note>
  You need permission to install apps in your Slack workspace, or a workspace admin who approves it.
</Note>

## Add your agent to Slack

<Steps>
  <Step title="Start the connection">
    Open your agent, click **Channels**, and click **Connect** on **Slack**. Slack's authorization page opens.
  </Step>

  <Step title="Allow access">
    Pick your workspace at the top right, choose the channel to post in, and click **Allow**.
  </Step>

  <Step title="Ask it something">
    In Slack, find the app under **Apps** in the sidebar and send it a message. To use it in a channel, add the app to the channel and @mention it with your question.
  </Step>
</Steps>

## How it works

* In direct messages the agent answers every message. In channels it answers when it's @mentioned, in a thread.
* Each person's conversation keeps its history. Type `/clear` to start over.
* Your [actions](/docs/agents/actions) work in Slack.

Once connected, open the Slack card to see **Manage Slack Workspaces**: each workspace with its channel, its status, and a switch to pause it. **Connect a New Slack Workspace** adds another workspace.

## Tips

* Train the agent on your internal documents: handbook, IT guides, benefits, product specs.
* Create a channel such as `#ask-the-agent` and add the app there, so answers are visible to everyone.
* In the [instructions](/docs/agents/instructions-and-model#instructions), say where to send people when the agent doesn't know, for example "Point people to #it-helpdesk."

<Columns>
  <Card title="Knowledge" icon="book" href="/docs/agents/knowledge">
    Upload your internal documents.
  </Card>

  <Card title="Channels" icon="share-nodes" href="/docs/agents/channels">
    All the places your agent can answer.
  </Card>
</Columns>


# Telegram
Source: https://easy-peasy.ai/docs/agents/telegram

Connect your agent to a Telegram bot so it answers people who message it on Telegram.

Your agent can answer on Telegram through a bot account you create with Telegram's @BotFather. People find your bot by its username and chat with it like with anyone else. Telegram is available on every plan, including Free.

## Connect Telegram

<Steps>
  <Step title="Create a Telegram bot">
    In Telegram, open a chat with **@BotFather** and send `/newbot`. Choose a display name, such as "Example Yoga Studio", and a username that ends in `bot`, such as `ExampleYogaBot`. BotFather replies with a token like `1234567890:AAF...`. Copy it and keep it private: anyone with the token controls your bot.
  </Step>

  <Step title="Paste the token">
    In Easy-Peasy.AI, open your agent, click **Channels**, and click **Connect** on **Telegram**. Paste the token into **Bot token** and click **Save**.

    <Frame>
      <img alt="The Telegram Integration window with a Bot token field, a Save button, and the Enable integration switch" />
    </Frame>
  </Step>

  <Step title="Say hello">
    In Telegram, search for your bot's username, send `/start`, and ask a question.
  </Step>
</Steps>

Each token can be connected to one agent. To put another agent on Telegram, create another bot with BotFather.

## How it works

* The bot answers in private chats. Each person's conversation keeps its history, so follow-up questions work.
* People can send photos and screenshots when **Allow image attachments** is on in [Appearance](/docs/agents/appearance#image-attachments). Other files aren't read.
* Your [actions](/docs/agents/actions) work on Telegram.
* Conversations appear in [**Conversations**](/docs/agents/conversations) under **Telegram**.

The bot understands three commands:

| Command | What it does |
| - | - |
| `/start` | Starts the conversation with a greeting |
| `/clear` | Clears the conversation history |
| `/ping` | Checks the bot is running; it answers "pong" |

## Make your bot look right

In @BotFather, `/setuserpic` sets the bot's profile photo and `/setdescription` sets the text people see before they tap **Start**. Share your bot's link, `t.me/YourBotUsername`, on your website and social profiles.

## Pause or change the token

To stop replies, open the Telegram card and switch off **Enable integration**. If your token leaks, send `/revoke` to @BotFather and connect the new token.

<Columns>
  <Card title="Channels" icon="share-nodes" href="/docs/agents/channels">
    All the places your agent can answer.
  </Card>

  <Card title="Conversations" icon="comments" href="/docs/agents/conversations">
    Read your Telegram chats.
  </Card>
</Columns>


# Voice calls
Source: https://easy-peasy.ai/docs/agents/voice

Let visitors talk to your agent out loud from the website chat, with the voice, language, and call limits you choose.

With voice calls, visitors tap **Talk** in the chat and talk to your agent. The agent listens, answers in the voice you pick, and uses the same knowledge and actions as in text chat. Visitors can still type during a call, for names, emails, or links.

For a phone number people can dial, see [Phone calls](/docs/agents/phone-calls).

<Note>
  Website visitors can call your agent on the Unlimited 200 and higher plans, Teams 200 and Teams 300, and Enterprise. On other plans, including Free, you can try voice calls in the Playground: 10 minutes every 30 days across all your agents, up to 3 minutes per call.
</Note>

## Turn on voice calls

<Steps>
  <Step title="Open Voice">
    Open your agent and click **Voice**. You can also use the **Voice** tab in the [Playground](/docs/agents/playground#set-up-voice-calls), which saves automatically.

    <Frame>
      <img alt="The Voice page: a banner saying voice calls on your website are available from Unlimited 200, the Enable voice calls switch on, a note that spoken replies use words at 2x the text rate, and Agent voice set to Hope, English, American, with a speaking speed of 1.00x" />
    </Frame>
  </Step>

  <Step title="Turn on Enable voice calls">
    A **Talk** button appears next to the message box. It turns into the send button as soon as the visitor types, and stays hidden while the agent is writing a reply.

    **Talk** is in English. If your visitors speak another language, write your own word in **Call button text**.
  </Step>

  <Step title="Pick a voice">
    Under **Agent voice**, click **Change**, play a few voices, and click **Select**.

    <Frame>
      <img alt="The Agent voice window with the tabs Voice library, My voices, and More voices, a search box, and voices such as Hope, Heather Rey, Matthew - Smooth, and Pro Narrator, each with a play button" />
    </Frame>

    | Tab | Voices |
    | - | - |
    | **Voice library** | About 1,900 voices in many languages and accents |
    | **My voices** | Voices you cloned with an Instant clone in [text to speech](/docs/audio/voices#instant-clone). Professional clones can't be used for calls |
    | **More voices** | Other voice families, with natural accents in several languages |
  </Step>

  <Step title="Click Save">
    Test a call in the Playground. Reload any open preview of your website chat to see the **Talk** button there.
  </Step>
</Steps>

## Settings

| Setting | What it does |
| - | - |
| **Call button text** | The word on the call button next to the message box, up to 16 characters. Empty shows **Talk** |
| **Allow text input during calls** | Keeps the message box usable during a call |
| **Allow interruptions** | Lets callers talk over the agent to cut it off |
| **Voice model** | The model used on calls. **Same as chat** uses the agent's model; the others are fast models, because every second the model thinks is a second of silence |
| **Sync with chat instructions** | Uses the agent's instructions on calls. Turn it off to write separate **Voice instructions**, for example "Keep answers to two sentences" |
| **Spoken language** | The language callers speak. **Multilingual** detects it; choosing the language makes recognition more accurate |
| **Silence before the agent answers** | How long a pause means the caller has finished. Shorter feels snappier; longer tolerates pauses |
| **Noise sensitivity** | How easily background noise counts as speech |
| **Speaking speed** | Slower or faster than the voice as recorded, for voices that support it |
| **Greeting** | What the agent says first. Empty reuses the chat's greeting |
| **Error message** | What the agent says when it doesn't catch something |
| **End the call after silence (seconds)** | Ends the call after this much silence. `0` keeps the line open until the caller hangs up |
| **Maximum call length (minutes)** | The longest a call can last, from 1 to 60 minutes |
| **Calls per visitor per day** | How many calls one visitor can make each day. It doesn't apply in the Playground |

## What calls cost

Spoken replies use your plan's words at twice the rate of a text reply, on every model, including models that don't use words in text chat. Listening, the voice, and the call itself are included.

Calls appear in [**Conversations**](/docs/agents/conversations) under **Web**, with **Voice call** as the source.

<AccordionGroup>
  <Accordion title="There's no Talk button in the chat">
    Check that **Enable voice calls** is on and saved, then reload the page with the chat. On plans below Unlimited 200, the **Talk** button appears only in the Playground.
  </Accordion>

  <Accordion title="The browser didn't ask for the microphone">
    Calls need microphone access. Allow it in the browser's site settings and try again. If you embed the chat with an iframe, keep `allow="microphone"` on it.
  </Accordion>

  <Accordion title="The agent answers too slowly">
    Pick a faster **Voice model**, shorten the **Voice instructions**, and lower **Silence before the agent answers**.
  </Accordion>
</AccordionGroup>

<Columns>
  <Card title="Phone calls" icon="phone" href="/docs/agents/phone-calls">
    Give your agent a phone number with Twilio.
  </Card>

  <Card title="Appearance" icon="palette" href="/docs/agents/appearance">
    Voice input for typed messages is under Chat features.
  </Card>
</Columns>


# Add your agent to your website
Source: https://easy-peasy.ai/docs/agents/website

Put a chat bubble on every page of your website with one line of code, embed the chat inside a page, or share a link to a hosted chat page.

There are three ways to put your agent on the web:

* **Chat bubble**: a button in the corner of every page that opens the chat. The most common choice.
* **Shareable link**: a hosted chat page. No website needed; share it in emails, QR codes, or your social profiles.
* **Iframe embed**: the chat inside a page, such as a help center or a contact page.

Open your agent, click **Channels**, and click **Manage** on **Chat bubble**. The **Website** page has all three.

## Add the chat bubble

<Steps>
  <Step title="Style the chat bubble">
    Click **Customize chat bubble**. It opens [**Appearance** > **Chat bubble**](/docs/agents/appearance#chat-bubble) in the Playground, where you choose where the bubble sits, its size, icon, color, and text. The **Website** page shows your saved settings and a preview of them on a sample page.

    **Open full preview** shows the real chat on a sample page.
  </Step>

  <Step title="Copy the install snippet">
    Under **Install snippet**, click **Copy**. The code is the same for every look, because the bubble's settings are saved with your agent:

    ```html theme={null}
    <script
        src="https://bots.easy-peasy.ai/chat.min.js"
        data-chat-url="https://bots.easy-peasy.ai/bot/YOUR_AGENT_ID"
        defer>
    </script>
    ```
  </Step>

  <Step title="Paste it into your website">
    Paste the code just before the closing `</body>` tag of your site, on every page where the chat should appear. Most site builders have a place for this, such as **Footer code** or **Custom code**: see [Website platforms](/docs/agents/website-platforms).
  </Step>

  <Step title="Check your site">
    Open your website. The chat bubble appears in the corner, and the greeting pops up next to it. Click the bubble and ask a question.

    <Frame>
      <img alt="A sample web page with the agent's chat bubble at the bottom right and its greeting, Welcome to Example Yoga Studio! How can I help you?, next to it" />
    </Frame>
  </Step>
</Steps>

Everything about the bubble and the chat, from its position and icon to the greeting and colors, is saved with your agent and updates on your site within about a minute, without changing the code. An older install code that still sets a position, size, color, or icon keeps working; what you save in [**Appearance**](/docs/agents/appearance#chat-bubble) takes priority over it.

<Frame>
  <img alt="Install snippet with a Copy button, Shareable link with the agent's hosted chat URL, and Iframe embed with an iframe code" />
</Frame>

## Shareable link

**Shareable link** is a chat page hosted by Easy-Peasy.AI, at `https://bots.easy-peasy.ai/bot/YOUR_AGENT_ID`. Open it or copy it with the buttons next to it. Share it anywhere you'd share a link, or turn it into a QR code.

## Iframe embed

**Iframe embed** puts the chat inside a page instead of floating over it. Copy the code and paste it where the chat should appear. To make it fill its space on every screen size, set the width to 100%:

```html theme={null}
<iframe
  src="https://bots.easy-peasy.ai/bot/YOUR_AGENT_ID?mode=embedded"
  width="100%"
  height="600"
  allow="microphone">
</iframe>
```

Keep `allow="microphone"` so visitors can use voice input and voice calls.

## Your Agent ID

**Your Agent ID** identifies your agent in embed codes, the [WordPress plugin](/docs/agents/website-platforms#wordpress), [Zapier](/docs/agents/integrations#zapier), and the [REST API](/docs/agents/integrations#rest-api).

## Control the chat from your website's code

Developers can open, close, and hide the chat from their own code, and pass the signed-in visitor's name and email to the agent. See [Chat widget for developers](/docs/developers/chat-widget).

<AccordionGroup>
  <Accordion title="The chat bubble doesn't appear">
    * Check that the code is before `</body>` and that you published your site.
    * In [**Settings**](/docs/agents/settings), **Is agent publicly available?** must be on.
    * If you set **Allowed Domains**, your website's domain must be in the list.
    * If your site uses a Content Security Policy, allow `https://bots.easy-peasy.ai` for scripts, frames, and connections (`connect-src`). Without `connect-src`, the bubble shows its default look instead of the one you saved.
  </Accordion>

  <Accordion title="The chat opens but says it can't answer">
    The agent may be paused because your plan's allowance ran out. See [Limits and usage](/docs/agents/limits-and-usage#when-an-allowance-runs-out).
  </Accordion>

  <Accordion title="The greeting pop-up stopped showing">
    After a visitor closes the pop-up or opens the chat, it stays hidden for that visitor for 24 hours. To change when it shows, or turn it off, use **Auto show initial messages pop-ups after (in seconds)** on [**Appearance**](/docs/agents/appearance).
  </Accordion>
</AccordionGroup>

<Columns>
  <Card title="Website platforms" icon="wordpress" href="/docs/agents/website-platforms">
    Steps for WordPress, Wix, Squarespace, and Webflow.
  </Card>

  <Card title="Appearance" icon="palette" href="/docs/agents/appearance">
    Name, greeting, and colors inside the chat.
  </Card>
</Columns>


# Website platforms
Source: https://easy-peasy.ai/docs/agents/website-platforms

Add your agent's chat to WordPress, Wix, Squarespace, Webflow, and other website builders.

Every platform needs the same thing: your agent's install snippet, pasted where the platform adds code before the closing `</body>` tag. Copy the snippet first from **Channels** > **Chat bubble** > **Manage** > **Install snippet**. See [Add your agent to your website](/docs/agents/website).

For a chat inside a single page instead of a floating bubble, use the **Iframe embed** code in the platform's HTML or embed block.

## WordPress

The **Easy-Peasy.AI Chatbot** plugin adds the chat bubble to every page without editing your theme.

<Steps>
  <Step title="Install the plugin">
    In your WordPress admin, go to **Plugins** > **Add New**, search for **Easy-Peasy.AI Chatbot**, click **Install Now**, and then **Activate**.
  </Step>

  <Step title="Copy your Agent ID">
    In Easy-Peasy.AI, open your agent, click **Channels** > **Chat bubble** > **Manage**, and copy **Your Agent ID**.
  </Step>

  <Step title="Paste it into the plugin">
    In WordPress, go to **Settings** > **Easy-Peasy.AI Bot** and paste the ID into **Bot ID**. Choose the **Widget Position**, **Button Color**, and an optional **Widget Icon URL**, then click **Save Settings**.
  </Step>
</Steps>

The plugin offers the four corner positions. For a side tab, a larger button, or a wider chat window, paste the install snippet instead, for example with a code snippets plugin or in your theme's footer. For a chat inside a page, add a **Custom HTML** block with the iframe code.

WordPress.com sites need a plan that allows plugins; on other plans, use the **Custom HTML** block with the iframe code.

## Wix

<Steps>
  <Step title="Open Custom code">
    In your Wix site's dashboard, go to **Settings** and open **Custom code** (search for "custom code" in the settings).
  </Step>

  <Step title="Add the snippet">
    Click **Add Code**, paste the install snippet, and give it a name. Under **Add Code to Pages**, choose **All pages** and **Load code once**. Set **Place Code in** to **Body - end**, then click **Apply**.
  </Step>

  <Step title="Check your live site">
    Open your published site. The chat bubble appears on every page.
  </Step>
</Steps>

If you don't see **Custom code**, your Wix plan may not allow it. Use the Wix Editor instead: **Add Elements** > **Embed Code** > **Embed HTML**, paste the iframe code, and publish.

## Squarespace

<Steps>
  <Step title="Open Code Injection">
    In your Squarespace dashboard, open **Code Injection** (search for "code injection" in the settings).
  </Step>

  <Step title="Paste the snippet into Footer">
    Paste the install snippet into the **Footer** field and click **Save**.
  </Step>

  <Step title="Check your live site">
    Open your site. The chat bubble appears on every page.
  </Step>
</Steps>

For a chat inside one page, edit the page, add a **Code** block, and paste the iframe code. Squarespace allows code injection and scripts only on some plans; if you don't see the option, check your plan.

## Webflow

<Steps>
  <Step title="Open Custom code">
    Open your site's **Site settings** and the **Custom code** tab.
  </Step>

  <Step title="Paste the snippet into Footer code">
    Paste the install snippet into **Footer code** and click **Save changes**.
  </Step>

  <Step title="Publish">
    Custom code goes live only when you publish. Click **Publish**, then open your live site.
  </Step>
</Steps>

To add the chat to one page only, paste the snippet into that page's **Before `</body>` tag** setting, or drag an **Embed** element onto the page and paste the iframe code. Webflow allows custom code on paid site plans.

## Shopify

Shopify has its own channel, which also lets the agent search your products and look up orders. See [Shopify](/docs/agents/shopify).

## Other websites

For any other site, paste the install snippet before `</body>` in your page template, or in your builder's footer or custom code setting. On single-page apps (React, Vue, Next.js), add the snippet once to the main HTML template or layout so it loads on every page.

<Columns>
  <Card title="Add your agent to your website" icon="window-maximize" href="/docs/agents/website">
    Style the bubble and copy the snippet.
  </Card>

  <Card title="Chat widget for developers" icon="code" href="/docs/developers/chat-widget">
    Open and hide the chat from your code.
  </Card>
</Columns>


# WhatsApp
Source: https://easy-peasy.ai/docs/agents/whatsapp

Connect a WhatsApp Business number to your agent in one click, keep using the WhatsApp Business app if you like, and manage message templates.

Your agent can answer messages sent to your WhatsApp Business number. You connect it by signing in with Facebook and picking or creating a number; there's no Meta developer app, token, or webhook to set up.

<Note>
  WhatsApp is available on paid plans. Meta charges its own fees for business messaging, billed to your Meta account.
</Note>

## Before you start

You need:

* A Facebook account.
* A phone number for your business that can receive a text message or a call for verification.
* A number that isn't signed in to the regular WhatsApp app. A number you use in the **WhatsApp Business** app is fine: see [Keep using the WhatsApp Business app](#keep-using-the-whatsapp-business-app).

## Connect in one click

<Steps>
  <Step title="Open the WhatsApp channel">
    Open your agent, click **Channels**, and click **Connect** on **WhatsApp**.

    <Frame>
      <img alt="The WhatsApp Integration window: Connect in one click, with a Continue with Facebook button and a link to enter credentials manually" />
    </Frame>
  </Step>

  <Step title="Choose how to connect">
    Click **Continue with Facebook**, then choose **Continue with Facebook** again for a number that isn't on WhatsApp yet, or **I already use the WhatsApp Business app**.

    <Frame>
      <img alt="Connect WhatsApp: Continue with Facebook, or I already use the WhatsApp Business app to keep using the app with messages working in both places" />
    </Frame>
  </Step>

  <Step title="Finish in Meta's window">
    Meta's window opens. Sign in with Facebook, pick or create a business portfolio, create your WhatsApp Business profile (the name customers see), add your number, and enter the code Meta sends. Keep the window open until it closes on its own.
  </Step>

  <Step title="Test it">
    The WhatsApp card shows **Connected**. Send a WhatsApp message to your business number from another phone; the agent replies within seconds.
  </Step>
</Steps>

Everything after Meta's window is automatic: your number is registered and incoming messages are sent to your agent.

## Keep using the WhatsApp Business app

Choose **I already use the WhatsApp Business app** to connect the number you already use in the app. The agent answers automatically, and you can still read and reply from the app on your phone. Your chat history stays.

Meta's requirements for this option:

* The number must be on the **WhatsApp Business** app, not regular WhatsApp, and in use there for at least 7 days on the latest version.
* It must not already be added to your WhatsApp Business account as an API number. If you tried the one-click option first, remove the number in WhatsApp Manager (**Phone numbers**), wait a few minutes, and try again.
* It isn't available in every country.

You'll confirm the connection in the app on your phone, so keep it nearby.

## The 24-hour rule and templates

This is WhatsApp's rule. Your agent can reply freely within 24 hours of a customer's last message, which covers normal back-and-forth. To start a conversation with someone who hasn't messaged you in the last 24 hours, WhatsApp requires an approved **message template**.

Once connected, click **Manage templates** in the WhatsApp window, then **New template**. Enter a **Template name**, **Category** (**Utility** or **Marketing**), the **Message body** with `{{1}}`-style variables and example values, and a **Language**, then click **Submit for approval**. Meta usually reviews templates within minutes to hours; the list shows each template's status.

## Pause, disconnect, or move a number

* **Pause:** switch off **Enable integration**. Messages still arrive at your number; the agent stops answering.
* **Disconnect:** click **Disconnect**. The number stays in your Meta account; it just stops going to this agent.
* **Move to another agent:** disconnect it here, then connect it on the other agent.

Conversations appear in [**Conversations**](/docs/agents/conversations) under **WhatsApp**. Your [actions](/docs/agents/actions) work on WhatsApp. The agent answers text messages; photos, voice notes, and documents aren't answered.

## Use your own Meta app

If your business already runs the WhatsApp Cloud API with its own Meta app, click **Using your own Meta app? Enter credentials manually** and fill in:

| Field | Where to find it |
| - | - |
| **WhatsApp phone number ID** | In your Meta app under **WhatsApp** > **API Setup** |
| **Access token** | A permanent system user token with the `whatsapp_business_messaging` and `whatsapp_business_management` permissions |
| **Verify token** | Any text you choose; enter the same text in Meta |
| **Webhook URL** | Shown in the form. In your Meta app, go to **WhatsApp** > **Configuration**, paste the URL and your verify token, and subscribe to the `messages` field |

Unless you need your own app, the one-click connection is faster and needs no upkeep.

<AccordionGroup>
  <Accordion title="The number is still in use in the WhatsApp app on a phone">
    The number is signed in to the regular WhatsApp app, and WhatsApp won't connect it while that account exists. Delete the WhatsApp account on the phone (**Settings** > **Account** > **Delete my account**), wait a few minutes, and connect again. If the phone uses WhatsApp Business, use [Keep using the WhatsApp Business app](#keep-using-the-whatsapp-business-app) instead.
  </Accordion>

  <Accordion title="It asks for a two-step verification PIN">
    The number was registered before and has a PIN. Enter it when asked, or turn off two-step verification in WhatsApp Manager (**Phone numbers** > **Settings** > **Two-step verification**) and connect again.
  </Accordion>

  <Accordion title="The connection expired before I finished">
    Meta allows 14 days to finish registering a number. Start the connection again from the beginning.
  </Accordion>

  <Accordion title="The agent isn't replying">
    Check that **Enable integration** is on, that the number shows as connected in WhatsApp Manager, and that you sent a text message.
  </Accordion>
</AccordionGroup>

<Columns>
  <Card title="Channels" icon="share-nodes" href="/docs/agents/channels">
    All the places your agent can answer.
  </Card>

  <Card title="Conversations" icon="comments" href="/docs/agents/conversations">
    Read your WhatsApp chats.
  </Card>
</Columns>


# Edit images
Source: https://easy-peasy.ai/docs/ai-images/edit

Change any image by describing the change or by pointing at spots on it, and keep every version.

You don't have to start over to change one thing. Edit an image you made, or upload your own, and describe what should be different.

## Open the editor

There are three ways in, and they all work the same:

* **From a result.** Hover over an image and click the pencil, or open it and click **Edit image**.
* **On its own page.** Choose **Images** > **Image Editor** from the menu at the top. Upload an image (JPG, PNG, or WebP) or pick one of your recent images.
* **From your own photo.** On AI Images, add a photo with **Upload**, **Character**, or **Object**. The **Action** menu appears, set to **Edit Image with AI**. Describe the change in **Prompt** and click **Generate**.

<Frame>
  <img alt="The AI Image Editor page: a box asking 'What do you want to change?', the AI Model menu set to GPT Image 2.5 Flare Medium, a Generate • 3 credits button, and a large preview of a yoga studio" />
</Frame>

## Describe the change

The editor has two ways to say what you want:

<Tabs>
  <Tab title="Prompt">
    Type the change in **What do you want to change?**, as if you were asking a designer. For example: "Make it evening and turn on the lamps" or "Change the mats to deep blue".
  </Tab>

  <Tab title="Visual">
    Click **Visual**, then click the spot on the image you want to change. A numbered marker appears; type what should happen there and click **Done**. Add as many markers as you need. This is the easiest way to change one object without touching the rest.

    <Frame>
      <img alt="Visual mode in the Edit Image window: marker 1 on a plant by the wall, with the instruction 'Replace this plant with a small wooden sign that says Sunrise Yoga'" />
    </Frame>
  </Tab>
</Tabs>

Then choose an **AI Model** and a size. **Auto** keeps the shape of your original. Click **Generate**; the button shows the price, for example **Generate • 3 credits**.

## Keep editing, version by version

Each edit adds a new version to the strip under the controls. Click **Original** or any version to see it, and keep editing from whichever one you like best. Every version is also saved in **History** > **Images**.

<Frame>
  <img alt="The edited yoga studio now has a wooden sign that reads Sunrise Yoga where the plant was, with the Original and the new version side by side in the version strip" />
</Frame>

## Add reference images

Click the paperclip to add other images the model should use, such as a product to place in the scene or a person to put in the picture. Only models that can combine several images are listed while references are attached: the FLUX.2, Seedream, Nano Banana, and GPT Image families.

## Which models can edit

Most image models can edit. These are available in the **AI Model** menu:

* **Every plan:** Nano Banana 2 Lite, Nano Banana 2, Grok, Muse Image, GPT Image 2 Low, GPT Image 2 Medium, GPT Image 2.5 Flare Low, GPT Image 2.5 Flare Medium, GPT Image 2.5 Sunburst Medium, Seedream 4.5, Seedream 5.0 Lite, FLUX.2 \[max], FLUX.2 \[pro], FLUX.2 \[flex], Seedream 4.0, Nano Banana, Qwen Image, Reve, Reve Fast, Mai Image 2.5, Flux Kontext Pro
* **Paid plans only:** Nano Banana Pro, Grok Pro, Grok Imagine 2.0, GPT Image 1.5, GPT Image 2 High, GPT Image 2.5 Flare High, GPT Image 2.5 Sunburst High, Seedream 5.0 Pro, Flux Kontext Max, GPT Image 1, Ideogram v3 Character

## Limits

* Upload JPG, PNG, or WebP files. On AI Images you can attach up to 8 photos at once. Very large images are scaled down to 3,840 pixels on the long side, and transparent areas become white.
* Uploads are checked against our content policy. If one is blocked, the **Generate** button changes to **Content Blocked**.
* There's no brush or mask. To change one area, use **Visual** markers.

<Columns>
  <Card title="Remove or replace a background" icon="scissors" href="/docs/ai-images/remove-background">
    Cut out the subject in one click.
  </Card>

  <Card title="Templates and one-click tools" icon="shapes" href="/docs/ai-images/templates-and-tools">
    Ready-made edits such as Ghiblify, Caricature, and mockups.
  </Card>
</Columns>


# Create images from text
Source: https://easy-peasy.ai/docs/ai-images/generate

Describe an image, add a style, camera, and lighting, pick a model, and generate.

You can go from an idea to a finished image in under a minute. Describe what you want, choose how it should look, and let one of the models draw it.

<Frame>
  <img alt="The left panel of AI Images with a prompt about a yoga studio at sunrise, the Customize buttons, the GPT Image 2.5 Flare Medium model, a 3:2 aspect ratio, and a Generate (3 credits) button" />
</Frame>

<Steps>
  <Step title="Describe the image">
    Type what you want to see in **Prompt**. Say what's in the picture, where it is, the light, and the style. For example: "A bright yoga studio at sunrise: wooden floor, rolled mats, tall windows with golden light, potted plants, calm and airy, photorealistic."

    Two buttons in the box can help. **Dictate** (the microphone) lets you speak your prompt. The button next to it writes a **Random prompt** when the box is empty, or rewrites yours in more detail (**Improve prompt**).
  </Step>

  <Step title="Add a style, camera, lighting, or shot (optional)">
    The **Customize** buttons add ready-made directions to your prompt:

    * **Style**: 43 looks, from **Oil Painting** and **Watercolor** to **Anime** and **Cyberpunk**.
    * **Camera**: presets such as **Cinematic** or **Phone Snap**, or a specific camera body, lens, and aperture.
    * **Lighting** and **Shot Type**: the light and the framing, such as **Golden Hour** light or a **Low Angle** shot.

    <Frame>
      <img alt="The Choose a style window with sample images for Oil Painting, Digital art, Detailed photo, Cyberpunk, Artstation, Watercolor, Pencil drawing, Concept art, Cartoon, and Anime" />
    </Frame>
  </Step>

  <Step title="Pick a model">
    Click the model under **Model** to open **Choose a model**. Every card says what the model is good at and what it costs per image. Search by name, or turn on **Multi-select** to generate the same prompt with several models at once and compare them. Compared images come out square, and the credits of each model add up.

    <Frame>
      <img alt="The Choose a model window with model cards for Nano Banana 2 Lite, Nano Banana 2, Nano Banana Pro, Muse Image, Grok Imagine 2.0, and Grok, each with a description and a credit price" />
    </Frame>

    The default is **GPT Image 2.5 Flare Medium**, a good all-rounder at 3 credits. See [Image models and credits](/docs/ai-images/models) for all of them.
  </Step>

  <Step title="Choose the size and how many">
    At the bottom of the panel, pick an aspect ratio, such as **1:1 Square**, **3:2 Standard**, or **16:9 Widescreen**, and how many images to make. You can make up to 4 at once, or 8 with some FLUX models. Models that offer it also show a **Resolution** setting, such as 1K, 2K, or 4K.
  </Step>

  <Step title="Click Generate">
    The button shows the total price, for example **Generate (3 credits)**. Your images appear in the **Images** tab on the right, usually within a minute.
  </Step>
</Steps>

## Put your own people, products, or photos in the picture

Use **Upload**, **Character**, or **Object** to add a photo. **Character** and **Object** keep a library of the people and products you add, so you can reuse them for consistent results across many images. You can also pick any of your past images from **History**.

When a photo is attached, the page switches to editing: an **Action** menu appears, set to **Edit Image with AI**, and your prompt describes what to make from the photo. See [Edit images](/docs/ai-images/edit).

**Photo Packs** turns a model you've trained on one person into themed photo sets, such as LinkedIn headshots or Halloween photos, from 3 credits per photo. It's available on paid plans.

## Tips for better images

* Be specific about the subject, the setting, and the light. "A golden retriever on a beach at sunset, low angle, warm light" beats "a dog".
* Put words you want to see in the image in quotes, and pick a model that's good at text, such as GPT Image 2.5 or Seedream 5.0 Pro.
* Try the same prompt with **Multi-select** to see which model suits it best, then stick with that one.
* To fix one detail, [edit the image](/docs/ai-images/edit) instead of generating again.

## On the Free plan

* One image at a time. Most wide and tall sizes are locked; with the default model, only the square **1:1** size is available.
* Models marked **Paid plans** in the picker are locked, and so are 4K, **Multi-select**, and Midjourney.
* Downloading needs a paid plan or bonus credits. See [Plans and credits](/docs/account/plans-and-credits#the-free-plan).

## If a generation fails or is blocked

* A tile that says **Generation Failed** doesn't cost anything. Try again, or try another model.
* If the GPT Image models refuse a prompt for policy reasons, we remake it with Seedream 5.0 Pro automatically. The tile then shows Seedream 5.0 Pro and is charged at that model's price.
* Uploaded photos are checked too. If one breaks the content policy, the button changes to **Content Blocked**.

<Columns>
  <Card title="Edit images" icon="pen" href="/docs/ai-images/edit">
    Change part of an image with words or markers.
  </Card>

  <Card title="Manage your images" icon="images" href="/docs/ai-images/manage">
    Download, share, upscale, or animate your results.
  </Card>
</Columns>


# Headshots
Source: https://easy-peasy.ai/docs/ai-images/headshots

Order a set of professional business headshots from a few selfies, for yourself or your whole team, with a one-time payment.

**Headshots** turns a few ordinary photos of you into a set of professional business headshots in different outfits, backgrounds, and lighting, for LinkedIn, your website, resumes, and company pages. It's a one-time purchase, separate from your plan, and available on every plan, including Free.

## Choose a package

Click **Images** > **Headshots** in the top menu, then **Get Your Headshots Now**.

<Frame>
  <img alt="Headshot pricing with an Individual and Team switch: Basic $19 one-time payment, 3 hours turnaround time, 60 headshots, 6 unique styles; Professional $29, marked Popular, 2 hours, 100 headshots, 15 unique styles; Executive $49, 1 hour, 180 headshots, 25 unique styles" />
</Frame>

| Package | Price | Headshots | Styles |
| - | - | - | - |
| **Basic** | \$19 | 60 | 6 |
| **Professional** | \$29 | 100 | 15 |
| **Executive** | \$49 | 180 | 25 |

Each purchase also adds 25 image & video credits to your account, for editing and upscaling your favourites.

To order for several people, switch to **Team**. See [Team headshots](#team-headshots).

## Order your headshots

<Steps>
  <Step title="Pay">
    Click the package's button and pay at checkout. After paying, you come back to upload your photos.
  </Step>

  <Step title="Upload 3 to 8 photos">
    Use recent photos of yourself alone, PNG or JPEG, at least 500 × 500 pixels, from different angles. Before you continue, a checklist asks you to confirm the photos meet the requirements:

    * no sunglasses, and only you in the photo,
    * not too close or too far from the camera,
    * looking at the camera, with a natural expression,
    * sharp, not blurry.

    <Warning>
      Photos that don't meet the requirements give poor results, and the order can't be refunded.
    </Warning>
  </Step>

  <Step title="Add your details">
    Enter your **Name**, **Age Range**, **Eye Color**, and **Gender**, and any **Additional Features**, such as glasses, a hijab, a turban, piercings, or your hair length. These help the headshots look like you.
  </Step>

  <Step title="Wait for your headshots">
    Your order shows **Generating headshots** on the Headshots page. It usually takes up to 2 hours; we email you when they're ready.
  </Step>
</Steps>

## Download and use your headshots

Open **Headshots** and click **View Results** on the order.

* **Download All Photos** downloads them all in a ZIP file.
* On each photo: **Edit image**, **Download image**, **Upscale image**, and **Generate Talking Video**.
* **Run the model in Image Generator** opens [AI Images](/docs/ai-images/photo-studio#create-photos-from-a-prompt) with you as the model, to create more photos of you from a prompt.

<Warning>
  Headshots are deleted automatically 30 days after they're ready. Download the ones you want to keep.
</Warning>

To order again, click **Create another headshot**. To delete an order and its photos, use its ⋯ menu.

## Team headshots

Order headshots for your whole team, with consistent styles, and pay once.

<Steps>
  <Step title="Choose the team size">
    Switch pricing to **Team** and set **Team Size**, from 3 to 50 people. Each person costs \$29, with a discount of 10% from 10 people, 20% from 20, and 30% from 30. The team size can't change after checkout.
  </Step>

  <Step title="Pay">
    Click **Get Headshots for Your Team**. After paying, click **Manage Team**.
  </Step>

  <Step title="Invite your team">
    On **Headshots Team**, click **Invite a member** and send the invitation by email or link. When someone joins, their order is ready for them to upload photos, and their status changes to **Model Created** once they have.
  </Step>
</Steps>

The owner sees each member's status, can start an order for a member, and can remove members. Members see their order marked **Team Shared**.

## What it costs

Headshots are a one-time payment in US dollars, charged at checkout. They don't use your plan's credits. Editing, upscaling, and talking videos made from your headshots use [image & video credits](/docs/account/plans-and-credits) as usual.

<Columns>
  <Card title="AI Photo Studio" icon="camera" href="/docs/ai-images/photo-studio">
    Create any photo of yourself from a prompt.
  </Card>

  <Card title="Talking videos" icon="comment" href="/docs/videos/talking-videos">
    Make a headshot speak.
  </Card>
</Columns>


# Interior designer
Source: https://easy-peasy.ai/docs/ai-images/interior-designer

Redesign a room from a photo, stage an empty space, change paint colors, remove clutter, or design a room from scratch.

**Interior Designer** reimagines rooms, gardens, and building exteriors. Upload a photo and say what to change, try it in a new style, furnish an empty room, or design a space from nothing. It's used by homeowners, real estate agents, designers, and anyone planning a makeover.

## Choose a mode

Click **AI Interior Designer** in the sidebar, or **Images** > **Interior Designer** in the top menu. Pick what to do under **Choose mode**:

<Frame>
  <img alt="The Choose mode menu: Edit, perfect for changing colors, materials, or adding and removing specific items; Restyle, choose from various design themes to transform your existing space; Generate, start from scratch with AI-generated room designs; 360° Panorama; and Declutter Room" />
</Frame>

| Mode | What it does |
| - | - |
| **Edit** | Changes what you describe: colors, materials, furniture, decor |
| **Restyle** | Redesigns your room in one or more design themes |
| **Generate** | Designs a room from scratch, without a photo |
| **360° Panorama** | Makes a 360° image of a room you can look around in |
| **Declutter Room** | Removes clutter and loose items |
| **Virtual Staging** | Furnishes an empty room |
| **Empty Room** | Removes the furniture |
| **Color / Paint Changer** | Repaints the walls in the color you describe |

## Edit a room

<Steps>
  <Step title="Upload a photo">
    Drag in a photo of the room, garden, or exterior: JPG, PNG, or WebP, up to 10 MB. For **Edit** and **360° Panorama**, **Add more** adds up to 6 photos of the same space.
  </Step>

  <Step title="Describe the change">
    With **Edit Mode** on **Prompt**, write it in **What would you like to edit?** Say where things go, for example "add a blue sofa on the left wall". To point at a spot instead, switch to **Visual** and click **Add Visual Markers**.

    <Frame>
      <img alt="Interior Designer with a photo of an empty studio with mirrors and a barre, Add Visual Markers, Add more, mode Edit, Edit Mode Prompt, the instruction to add rolled yoga mats and round bolsters on low wooden shelves and a few sage green floor cushions, and Output Settings 16:9 and 1K" />
    </Frame>
  </Step>

  <Step title="Choose the output settings">
    Pick the aspect ratio and the resolution: **1K**, **2K**, or **4K**. 4K is for paid plans.
  </Step>

  <Step title="Click Generate">
    The result appears on the right in about 20 seconds.

    <Frame>
      <img alt="The result: the same studio with low wooden shelves of rolled mats and bolsters along the back wall and sage green floor cushions, with Video, Upscale, Edit, and Download buttons, and the Original and the result below to compare" />
    </Frame>
  </Step>
</Steps>

Click the thumbnails under the result to compare it with the original. To keep changing the result, describe the next change and click **Generate** again. **Start Fresh** begins again with a new photo.

Each result has:

* **Edit**: opens it in an editing window to change it with a description, like [editing images](/docs/ai-images/edit).
* **Upscale**: opens it in the Magic Upscaler in a new tab. See [Upscale images](/docs/ai-images/upscale).
* **Video**: turns it into a short video, such as a slow walkthrough. See [Generate videos](/docs/videos/generate).
* **Download**, and thumbs up and down.

Your results are also saved in [AI Images](/docs/ai-images/manage).

## Restyle or design from scratch

For **Restyle** and **Generate**, choose the **Room Type**, such as a living room, kitchen, home office, coffee shop, or front yard, and one or more **Themes**: Minimalist, Scandinavian, Industrial, Bohemian, Japanese Zen, Coastal, and many more. **Custom** lets you describe your own. You get one design per theme, so picking three themes makes three designs.

<Frame>
  <img alt="A generated minimalist Pilates and barre studio with mirrors, a wooden barre, light oak floors, and floor-to-ceiling windows, labeled Minimalist, with Video, Upscale, Edit, and Download" />
</Frame>

In **Generate**, you also pick the **Model** and the aspect ratio. Add details in **Additional Prompt (Optional)**, such as colors or furniture you want.

## Tips

* Take the photo in daylight, from a corner, so the whole room is in view.
* In Edit, change one or two things at a time, then build on the result.
* For paint colors, name the color and the finish, such as "soft sage green, matte".

## What it costs

Interior Designer uses [image & video credits](/docs/account/plans-and-credits):

| Mode | Credits per image |
| - | - |
| **Edit**, **Restyle**, **Declutter Room**, **Virtual Staging**, **Empty Room**, **Color / Paint Changer** | 7 at 1K or 2K, 12 at 4K |
| **Generate** | Depends on the model: 4 with the default, Nano Banana 2 |
| **360° Panorama** | 3, or 11 with your own photos |

<Note>
  For the photo modes, the **Generate** button currently shows a lower number than what's charged. The table above is what's used.
</Note>

On the Free plan, you can pick one theme at a time, and the photo modes need more credits than the Free plan's 3 a month.

<Columns>
  <Card title="Edit images" icon="pen-to-square" href="/docs/ai-images/edit">
    Change any image with a description.
  </Card>

  <Card title="Generate videos" icon="video" href="/docs/videos/generate">
    Turn a room into a walkthrough video.
  </Card>
</Columns>


# Manage your images
Source: https://easy-peasy.ai/docs/ai-images/manage

Find, download, share, animate, and delete your images, and get credits back for a result you're not happy with.

## Where your images are

* **The Images tab** on AI Images shows your latest results.
* **History** > **Images** has everything you've made, including videos. Search by prompt, filter by date, and on a team, filter by person.
* **History** > **Favorites** collects the images you've marked with the heart.

## Quick actions

Hover over an image to see its actions:

<Frame>
  <img alt="An image with its action bar: share, open, regenerate, edit, video, talking video, improve, and download, plus a heart in the corner" />
</Frame>

| Button | What it does |
| - | - |
| Share | Share the image publicly or with your team |
| Open | Open the image's details |
| Regenerate | Reopen AI Images with the same prompt, model, and settings |
| Edit Image | [Edit the image](/docs/ai-images/edit) |
| Generate Video | Turn the image into a video |
| Talking Video | Make the person in the image speak your script |
| Improve | [Upscale the image](/docs/ai-images/upscale) |
| Download | Save the image to your device |
| Heart | Add the image to your favorites |

## Image details

Click an image to see how it was made: the model, the prompt (with a button to copy it), the credits it used, and the date. The same actions are here as buttons, plus **Remove Background** and **Delete**.

<Frame>
  <img alt="The details of an image: model GPT Image 2.5 Flare Medium, the prompt, 3 credits used, a Share with Team switch, and buttons for Regenerate, Edit image, Upscale image, Generate video, Generate Talking Video, Download image, and Remove Background" />
</Frame>

## Share an image

* **Share with Team** in the details makes the image visible to everyone on your team. Teams plans only.
* **Share** gives you a public link, a page for the image on easy-peasy.ai, and buttons to post it to LinkedIn, Facebook, X, and Reddit. Images you share publicly can also appear in **Examples** for other people to see. Turn **Share Publicly** off to take the page down.

Images stay private unless you share them.

## Turn an image into a video

Click **Generate Video**, pick a video model and a length, and describe the motion. The video appears on the [AI Videos](/docs/videos/generate) page when it's ready. Video uses image & video credits and is available on paid plans.

## Get credits back for a bad result

If an image came out wrong, click the thumbs-down on it. You get its credits back, up to 10% of your plan's monthly credits in each billing period. Your feedback also helps us improve the models.

## Delete an image

Open the image and click **Delete**. Deleted images can't be restored, and deleting doesn't return the credits they used.

## Download

**Download** saves the image in its original format. On the Free plan, downloading needs bonus credits; see [Plans and credits](/docs/account/plans-and-credits#the-free-plan).


# Image models and credits
Source: https://easy-peasy.ai/docs/ai-images/models

Every image model on Easy-Peasy.AI, what it's good for, what it costs per image, and whether the Free plan can use it.

You choose a model in **Choose a model** on [AI Images](/docs/ai-images/generate). This page lists the same models, in the same order, with the same prices.

## How prices work

* **Per image.** The price is for one image. Making four images costs four times as much.
* **By resolution.** Some models charge more for bigger images, for example 4 credits at 1K and 8 credits at 4K. Pick the resolution under **Resolution**.
* **Shown before you start.** The **Generate** button shows the total for your settings.

The default model is **GPT Image 2.5 Flare Medium** at 3 credits per image.

## Text-to-image models

| Model | Good for | Credits per image | Free plan |
| - | - | - | - |
| Nano Banana 2 Lite | Google's Gemini 3.1 Flash-Lite Image model — the fastest, most cost-efficient Nano Banana for high-quality image generation and editing at \~4s per image. | 2 credits per image | Yes |
| Nano Banana 2 | Google's Gemini 3.1 Flash Image model for general image generation and editing with high quality and low latency. | 3 credits (0.5K), 4 credits (1K), 6 credits (2K), 8 credits (4K) | Yes |
| Nano Banana Pro | Google's Gemini 3 Pro model for advanced image generation with superior quality and detail. | 7 credits (1K/2K), 12 credits (4K) | No |
| Muse Image | Meta's Muse Image — an agentic image model that can look up visual references and current facts before it renders, with strong text, infographic and product accuracy. Generates and edits. | 1 credit | Yes |
| Grok Imagine 2.0 | xAI's latest Grok Imagine 2.0 model with improved photorealism, prompt adherence, aspect ratio control, and image editing support. | 3 credits (1K), 4 credits (2K) | Yes |
| Grok | xAI's Grok image generation model with unique stylistic qualities, aspect ratio control, and image editing support. | 1 credit | Yes |
| Grok Pro | xAI's premium Grok image generation model with superior quality, aspect ratio control, and image editing support. | 4 credits | No |
| GPT Image 2 Low | OpenAI's GPT Image 2 at low quality — fastest and cheapest, great for drafts, thumbnails, and quick iterations. | 0.5 credits per image | Yes |
| GPT Image 2 Medium | OpenAI's GPT Image 2 at medium quality — a balanced option for most production use cases. | 3 credits per image | Yes |
| GPT Image 2 High | OpenAI's GPT Image 2 model at high quality for maximum detail, sharpest text rendering, and best prompt adherence. | 11 credits per image | No |
| GPT Image 2.5 Flare Low | OpenAI's GPT Image 2.5 Flare at low quality — up to 50% faster than GPT Image 2, ideal for drafts, thumbnails, and rapid iteration. | 0.5 credits per image | Yes |
| GPT Image 2.5 Flare Medium | OpenAI's GPT Image 2.5 Flare at medium quality — OpenAI's recommended default: more natural lighting, richer textures, better reference-photo preservation, and up to 50% lower latency than GPT Image 2. | 3 credits per image | Yes |
| GPT Image 2.5 Flare High | OpenAI's GPT Image 2.5 Flare at high quality — maximum detail and the sharpest text rendering, still up to 50% faster than GPT Image 2 High. | 11 credits per image | No |
| GPT Image 2.5 Sunburst Medium | OpenAI's GPT Image 2.5 Sunburst at medium quality — the premium tier for tighter control across multi-step edits, campaign creative, and polished product imagery (slower than Flare). | 3 credits per image | Yes |
| GPT Image 2.5 Sunburst High | OpenAI's GPT Image 2.5 Sunburst at high quality — production-ready campaign assets and product shots with the most precise editing control OpenAI offers. | 11 credits per image | No |
| GPT Image 1.5 | Most advanced OpenAI image model with improved image preservation, better instruction following, enhanced text rendering, and 20% lower cost. | 2 credits per image (standard), 8 credits (high quality) | Yes |
| Seedream 4.5 | ByteDance's new-generation image creation model with superior quality and prompt adherence. | 2 credits per image | Yes |
| Seedream 5.0 Lite | ByteDance's fast Seedream 5.0 Lite model with high quality intelligent text-to-image generation. | 2 credits per image | Yes |
| Seedream 5.0 Pro | ByteDance's flagship Seedream 5.0 Pro model with deep-thinking prompt understanding, native text in 14 languages, and precise control over dense layouts. | 4 credits (1K), 8 credits (2K) | No |
| Qwen Image Max | Alibaba's flagship Qwen image model with superior prompt understanding, dense text rendering, and rich detail. | 4 credits per image | No |
| Qwen Image 2.0 | Alibaba's new-generation Qwen image model with excellent prompt following in Chinese and English and strong text rendering. | 2 credits per image | Yes |
| WAN 2.7 Pro | Alibaba's premium WAN 2.7 image model with deep prompt reasoning for the highest quality photorealistic output. | 4 credits per image | No |
| WAN 2.7 | Alibaba's WAN 2.7 image model with prompt reasoning, ultra-realistic textures, and excellent prompt adherence. | 2 credits per image | Yes |
| Krea v2 Large | Krea's largest v2 model with distinctive aesthetics and adjustable creativity for striking, non-generic images. | 3 credits per image | Yes |
| Krea v2 Medium | Krea's balanced v2 model with great aesthetics and quality at a mid-tier price. | 2 credits per image | Yes |
| Krea v2 Medium Turbo | Krea's fastest v2 model — great aesthetics at high speed and low cost. | 1 credit per image | Yes |
| Luma Uni v1 | Luma's unified image model with strong composition, lighting, and support for wide and tall aspect ratios. | 2 credits per image | Yes |
| Boogu Image | High-quality text-to-image generation with strong aesthetics and reliable prompt adherence. | 3 credits per image | Yes |
| Cosmos 3 Super | NVIDIA's Cosmos 3 Super model for physically grounded, photorealistic image generation. | 2 credits per image | Yes |
| P-Image | Pruna's ultra-efficient image model — solid quality at the lowest cost and very high speed. | 0.5 credits per image | Yes |
| Z-Image Turbo | Super fast text-to-image model with excellent quality and speed. | 0.1 credits (1K), 1 credit (2K) | Yes |
| ImagineArt 1.5 | High-fidelity professional-grade visuals with lifelike realism, strong aesthetics, and accurate text rendering. | 2 credits per image | Yes |
| Mai Image 2.5 | Microsoft's photorealistic model with pixel-level control and strong typography, turning prompts into design-ready visuals. | 2 credits per image | Yes |
| FLUX.2 \[max] | Maximum quality AI image generation with superior detail and prompt adherence. Optimized for the highest quality production assets. | 2.5 credits (1K), 5 credits (2K) | Yes |
| FLUX.2 \[flex] | Production-grade AI image generation with 4MP photorealistic output. Highest detail and prompt adherence for final production assets. | 3 credits (1K), 6 credits (2K) | Yes |
| FLUX.2 \[pro] | Best balance of speed and cost. Ideal for high-volume applications needing fast turnaround. | 2 credits per image | Yes |
| FLUX.2 \[dev] | Enhanced realism, crisper text generation, and native editing capabilities. | 1 credit per image | Yes |
| HunyuanImage 3.0 | Best open-source text-to-image model with complex prompt understanding and text generation within images. | 5 credits per image | No |
| Nano Banana | Google's Gemini 2.5 Flash model for creative image generation with excellent prompt understanding. | 2 credits per image | Yes |
| REVE | Generates detailed visual output that closely follows your instructions, with strong aesthetic quality and accurate text rendering. | 2 credits per image | Yes |
| Flux.1 Krea | FLUX.1 Krea is a new model that generates high-quality images from text with incredible aesthetics. | 2 credits per image | Yes |
| Flux Kontext Pro | Optimized for detailed prompts and iterative editing with excellent character consistency; supports reference images. | 2 credits per image (standard), 4 credits (HD) | Yes |
| Flux Kontext Max | The most advanced model for high precision and superior prompt adherence; supports reference images, complex scene editing, and flawless typography. | 4 credits per image | No |
| GPT Image 1 | New OpenAI model with specific capabilities. Check OpenAI docs for details. | 2 credits per image (standard), 8 credits (high quality) | Yes |
| HiDream | High-quality dream-like imagery with excellent detail and composition. | 2 credits per image | Yes |
| WAN 2.2 | Exceptional prompt adherence and ultra realisic textures. | 2 credits per image | Yes |
| WAN 2.5 Preview | High-quality image generation with various artistic styles. Support for realistic photography and creative applications. | 2 credits per image | Yes |
| Bria 3.2 | Trained exclusively on licensed data for safe commercial use. Excels in text rendering and aesthetics. | 2 credits per image | Yes |
| MiniMax Image 01 | MiniMax's latest text-to-image model with excellent prompt following and high-quality visual generation. | 0.5 credits per image | Yes |
| Seedream 4.0 | ByteDance's advanced text-to-image model with excellent prompt following and visual quality. | 2 credits per image | Yes |
| Seedream 3.0 | ByteDance's advanced text-to-image model with excellent prompt following and visual quality. | 2 credits per image | Yes |
| Dreamina 3.1 | Dreamina 3.1 is a new model that generates high-quality images from text with incredible aesthetics. | 2 credits per image | Yes |
| Qwen-Image | Qwen-Image is a new model that generates high-quality images from text with incredible aesthetics. | 2 credits per image | Yes |
| Ideogram v4 | Ideogram's latest model for crisp visuals, accurate text rendering, and polished posters and logos. | 3 credits per image | Yes |
| Ideogram v3 | Best for prompt comprehension and text rendering. | 3 credits per image (standard), 5 credits (high quality) | Yes |
| Recraft v4.1 | Design-first model for brand systems and editorial work, delivering production-ready images with sharp prompt control and clean composition. | 2 credits per image | Yes |
| Recraft v4.1 Pro | The highest-quality Recraft v4.1 tier for raster images — maximum detail and prompt adherence for premium production assets. | 10 credits per image | No |
| Recraft v4.1 Vector | Turns prompts into fully editable SVGs with structured layers and clean geometry — ideal for logos, icons, and illustration systems. | 4 credits per image | No |
| Recraft v4.1 Pro Vector | The highest-quality Recraft v4.1 tier for editable SVGs — premium logos, icons, and illustration systems with the cleanest geometry. | 10 credits per image | No |
| Recraft v3 | Recommended for best results. 2 credits per image. | 2 credits per image | Yes |
| FLUX 1.1 Pro Ultra | Latest FLUX model with superior quality and raw mode for photorealistic images. | 4 credits per image | No |
| Midjourney V6 | Best for high-quality images. 2 credits per image. | 2 credits per image | Yes |
| Midjourney V7 | Best for high-quality images. 2 credits per image. | 2 credits per image | Yes |
| FLUX.1 | State-of-the-art image generation with top of the line prompt following, visual quality, image detail and output diversity. | 0.25 credits per image | Yes |
| FLUX 1.1 Pro | State-of-the-art image generation with superior prompt following, visual quality, and diversity, delivering better quality results faster. | 2 credits per image | Yes |
| Stable Diffusion 3.5 | Latest Stable Diffusion model with improved quality and speed | 2 credits per image | Yes |
| Stable Diffusion 3.0 | 2 credits per image | 2 credits per image | Yes |
| Stable Diffusion XL | | 0.25 credits per image | Yes |

On the Free plan, 4K resolution and Midjourney aren't available either, even where a model is marked Yes.

## Editing models

Editing usually costs the same as generating with that model. The **Generate** button in the editor shows the price.

* **Every plan:** Nano Banana 2 Lite, Nano Banana 2, Grok, Muse Image, GPT Image 2 Low, GPT Image 2 Medium, GPT Image 2.5 Flare Low, GPT Image 2.5 Flare Medium, GPT Image 2.5 Sunburst Medium, Seedream 4.5, Seedream 5.0 Lite, FLUX.2 \[max], FLUX.2 \[pro], FLUX.2 \[flex], Seedream 4.0, Nano Banana, Qwen Image, Reve, Reve Fast, Mai Image 2.5, Flux Kontext Pro
* **Paid plans only:** Nano Banana Pro, Grok Pro, Grok Imagine 2.0, GPT Image 1.5, GPT Image 2 High, GPT Image 2.5 Flare High, GPT Image 2.5 Sunburst High, Seedream 5.0 Pro, Flux Kontext Max, GPT Image 1, Ideogram v3 Character

## Your own models

Models you train in **AI Photo Studio** appear under **My Custom Models** in **Choose a model**. Use them to generate photos of the person you trained them on.


# AI Images
Source: https://easy-peasy.ai/docs/ai-images/overview

Create, edit, and improve images with dozens of AI models, templates, and one-click tools in one place.

<Frame>
  <img alt="The AI Images page: the prompt, Customize options, and Generate button on the left, and three finished images on the right" />
</Frame>

Open **AI Images** in the sidebar, or choose **Images** > **Image Generator** from the menu at the top of any tool page. The page has two halves:

* **On the left**, you describe the image, add your own photos if you want them in it, pick a model, and click **Generate**.
* **On the right**, you see your results and ideas to start from.

| Tab on the right | What's in it |
| - | - |
| **Images** | Your latest results. Everything older is in **History** > **Images**. |
| **Templates** | Ready-made looks you can reuse with your own photos, such as mockups, posters, and portrait styles. |
| **AI Image Tools** | One-click jobs such as **Remove Background**, **Ghiblify**, and **Caricature**. |
| **Examples** | Images other people chose to share publicly, for inspiration. |

## What you can do

<Columns>
  <Card title="Create images from text" icon="wand-magic-sparkles" href="/docs/ai-images/generate">
    Describe an image, choose a style and a model, and generate up to four at once.
  </Card>

  <Card title="Edit images" icon="pen" href="/docs/ai-images/edit">
    Change any image by describing the change, or by pointing at spots on it.
  </Card>

  <Card title="Templates and one-click tools" icon="shapes" href="/docs/ai-images/templates-and-tools">
    Start from a ready-made look, or run a tool like Ghiblify or Caricature.
  </Card>

  <Card title="Remove or replace a background" icon="scissors" href="/docs/ai-images/remove-background">
    Cut out the subject, or put it somewhere new.
  </Card>

  <Card title="Upscale images" icon="up-right-and-down-left-from-center" href="/docs/ai-images/upscale">
    Make an image sharper and up to four times larger, or restore an old photo.
  </Card>

  <Card title="Manage your images" icon="images" href="/docs/ai-images/manage">
    Download, share, turn into video, and find everything you've made.
  </Card>
</Columns>

## More image tools

The **Images** menu at the top of the page also opens these:

| Tool | What it does |
| - | - |
| **Image Editor** | A full-page version of the editor. See [Edit images](/docs/ai-images/edit). |
| **Magic Upscaler** | See [Upscale images](/docs/ai-images/upscale). |
| **Headshots** | Professional headshots of you or your team from a few selfies. |
| **AI Photo Studio** | Train your own model from photos of a person, then generate themed photo packs with it. |
| **Interior Designer** | Redesign a room from a photo. |
| **Remove Background** and **Replace Background** | See [Remove or replace a background](/docs/ai-images/remove-background). |

## Credits

Every image uses image & video credits. The price depends on the model and, for some models, the resolution. The **Generate** button shows the total before you start. See [Image models and credits](/docs/ai-images/models) for every model's price, and [Plans and credits](/docs/account/plans-and-credits) for how credits work.


# AI Photo Studio
Source: https://easy-peasy.ai/docs/ai-images/photo-studio

Upload a few photos of a person and create new photos of them in any setting, from a prompt or with ready-made photo packs.

**AI Photo Studio** makes new photos of a person from a few pictures of them: professional portraits, lifestyle shots, travel photos, or creative looks. You set up a **model** of the person once, then describe any scene with them in it.

Use it for profile pictures, social media, marketing with your team's faces, or fun portraits. Only upload photos of yourself, or of people who agreed to it.

<Note>
  Creating a model and photo packs are for paid plans.
</Note>

## Create a model

<Steps>
  <Step title="Open AI Photo Studio">
    Click **Images** > **AI Photo Studio** in the top menu, or go to [easy-peasy.ai/ai-images/models](https://easy-peasy.ai/ai-images/models). Click **Start Your AI Photo Shoot Now**, or **Train New Model** if you already have one.
  </Step>

  <Step title="Upload 3 to 10 photos">
    Drag in photos of the same person, PNG or JPEG. Use sharp, well-lit photos where only that person is in view, from different angles and in different settings. Avoid group photos and blurry or heavily filtered pictures.

    <Frame>
      <img alt="Upload 3-10 images of the character, with an upload area and three uploaded photos of the same curly-haired woman: a studio portrait, in a white shirt in a yoga studio, and laughing in a park, and a Continue button" />
    </Frame>
  </Step>

  <Step title="Name the model">
    Click **Continue**, then enter a **Name** and choose the **Gender**, and click **Continue**. The model is ready straight away.

    <Frame>
      <img alt="Add additional information with Name Maya and Gender Female, and Continue" />
    </Frame>
  </Step>

  <Step title="Pick photo packs, or skip them">
    Next you can choose [photo packs](#photo-packs). To create your own scenes instead, go back to **AI Photo Studio** and click **Run Model**.
  </Step>
</Steps>

## Create photos from a prompt

<Frame>
  <img alt="Your AI models with one model, Maya, created on 9/26/2026, a Run Model button and a menu, and Train New Model" />
</Frame>

Click **Run Model**. AI Images opens with your model selected under **Model**.

<Steps>
  <Step title="Describe the scene">
    Write the prompt with the word **Model** where the person should be, for example: "Model teaching a morning yoga class on a wooden deck by the sea, smiling, wide shot, photorealistic".
  </Step>

  <Step title="Pick a rendering model">
    **Rendering model** decides which image model draws the photo. **Hyperrealism** gives the most lifelike results.

    <Frame>
      <img alt="The Image Generator with the prompt Model teaching a morning yoga class on a wooden deck by the sea, the Customize options, Model set to Maya, Custom trained model, Rendering model Hyperrealism, and Generate (3 credits)" />
    </Frame>
  </Step>

  <Step title="Click Generate">
    The photos appear with your other images a few seconds later.

    <Frame>
      <img alt="A generated photo of the same woman sitting cross-legged on a wooden deck by the sea, leading a group yoga class" />
    </Frame>
  </Step>
</Steps>

| Rendering model | Credits per image |
| - | - |
| **Hyperrealism** (default) | 3 |
| **Nano Banana 2** | 4 |
| **Nano Banana Pro** | 7 |
| **Nano Banana 2 Lite** | 2 |
| **Seedream 5 Pro** | 8 |
| **Seedream 5 Lite** | 2 |
| **Legacy** | 2 |

Everything else in AI Images works as usual: aspect ratio, number of images, styles, and editing the results. See [Generate images](/docs/ai-images/generate).

## Photo packs

Photo packs are ready-made photo shoots: pick a pack, such as Halloween, fitness, or business portraits, and get a whole set of photos of your person in that theme, without writing prompts. Search the packs, or filter them by **Category** and **Gender**.

<Frame>
  <img alt="The photo packs bar: Back, 0 packs selected, Credits: 398 remaining, and Generate Photos" />
</Frame>

Each pack says how many photos it makes, about 30 to 100. Select one or more packs and click **Generate Photos**. The photos appear in AI Images.

Photo packs use 3 image & video credits per photo, so a 32-photo pack uses 96 credits. The bar at the bottom shows how many packs you've selected and the credits you have left.

## Manage your models

**AI Photo Studio** lists your models, including [headshot](/docs/ai-images/headshots) orders. To delete one, click the ⋯ menu next to **Run Model**, then **Delete**.

<Columns>
  <Card title="Headshots" icon="user-tie" href="/docs/ai-images/headshots">
    Order a set of professional headshots.
  </Card>

  <Card title="Generate images" icon="image" href="/docs/ai-images/generate">
    All the options in AI Images.
  </Card>
</Columns>


# Remove or replace a background
Source: https://easy-peasy.ai/docs/ai-images/remove-background

Cut the subject out of any image, or place it in front of a new background you describe.

## Remove a background

<Tabs>
  <Tab title="An image you made">
    Open the image from **Images** or **History** > **Images** and click **Remove Background**. The cut-out is saved as a new image in **History** > **Images**, and the original stays as it is. It costs 1 credit.

    <Frame>
      <img alt="The result of Remove Background: a laptop and desk cut out onto a transparent background, shown as black, with Credits used: 1" />
    </Frame>
  </Tab>

  <Tab title="Your own photo">
    1. Choose **Images** > **Remove Background** from the menu at the top.
    2. Drop your photo into **Upload Image**.
    3. Click **Generate**.

    It costs 1 credit, and the result appears in **Images**.
  </Tab>
</Tabs>

The result has a transparent background. Image viewers that don't support transparency show it as black or checkered, but it's transparent when you place it on a slide, website, or design.

## Replace a background

1. Choose **Images** > **Replace Background** from the menu at the top.
2. Upload your photo.
3. In **Prompt**, describe the new background, for example "a white marble countertop in a bright kitchen".
4. Click **Generate**.

The subject stays as it is and the background is replaced. It costs 2 credits per image.

<Tip>
  For product photos, describe the surface, the light, and the setting: "on a wooden table by a window, soft morning light, shallow depth of field". A specific background looks more natural than a vague one.
</Tip>

## Other ways to change a background

* The **Change Background** and **Product Background Change** templates do the same job with a tested prompt. See [Templates and one-click tools](/docs/ai-images/templates-and-tools).
* To change the background and something else at the same time, [edit the image](/docs/ai-images/edit) and describe both changes.


# Templates and one-click tools
Source: https://easy-peasy.ai/docs/ai-images/templates-and-tools

Start from a ready-made look, or run a one-click tool such as Ghiblify, Caricature, or Remove Background.

When you don't want to write a prompt from scratch, start from a template or a tool. Both live in the right half of **AI Images**.

## Use a template

Templates are ready-made looks: product mockups, posters, portrait styles, filters, and more. Each one comes with a tested prompt, the model it was made for, and a sample photo you swap for your own.

<Steps>
  <Step title="Find a template">
    Open the **Templates** tab and filter by category, such as **New**, **Starring You**, **Product**, or **Mockups**. Hover over a template and click **Use template**.
  </Step>

  <Step title="Swap in your photo">
    The left panel fills in: the template appears at the top, with its prompt and model below it. Its sample photo sits in the **Customize** row. Remove the sample with its X, then add your own photo with **Upload**, **Character**, or **Object**, or pick one of your past images from **History**.

    <Frame>
      <img alt="The Laptop Screen Mockup template loaded in the left panel, with its prompt, the yoga studio image added as image1 in the Customize row, and a Generate (3 credits) button" />
    </Frame>

    Templates in **Starring You** need a photo of a person. Product templates need a photo of the product.
  </Step>

  <Step title="Generate">
    Change the prompt if you like, then click **Generate**. This example put the yoga studio image on a laptop screen:

    <Frame>
      <img alt="A laptop on a wooden desk showing the yoga studio image on its screen, with books, a mug, and plants around it" />
    </Frame>
  </Step>
</Steps>

To stop using a template, click the X on the template card at the top of the panel. It clears what the template filled in and keeps the rest.

<Tip>
  When you switch from one template to another, check the **Customize** row for a photo left over from the previous template, and remove it if you don't need it.
</Tip>

## One-click tools

The **AI Image Tools** tab lists jobs that need only a photo. Some also appear in the **Action** menu that shows up when you add a photo on AI Images.

<Frame>
  <img alt="The AI Image Tools tab with cards for Train my AI Model, Visualize What Happens Next, Hyper Realistic Photos, Edit Image with AI, Turn photo into muppets, Ghiblify, and more" />
</Frame>

| Tool | What it does | Credits |
| - | - | - |
| **Remove Background** | Cuts the subject out onto a transparent background | 1 |
| **Replace Background** | Puts the subject in front of a new background you describe | 2 per image |
| **Caricature** | Turns a portrait into a caricature | 2 per image |
| **Ghiblify** | Redraws a photo as Studio Ghibli–style artwork | 2 per image |
| **Turn photo into muppets** | Turns the people in a photo into puppets | 2 per image |
| **Hyper Realistic Photos** | Makes a photo look more realistic and detailed | 2 per image |
| **Visualize What Happens Next** | Draws a 3×3 grid of scenes that continue the story of your image | 12 |
| **Stylization** | Makes eight stylized portraits at once, such as manga, claymation, oil painting, and sketch. Paid plans | 16 |
| **Colorize** (in the **Action** menu) | Adds color to a black-and-white photo | 8 per image |
| **Character Reference** (in the **Action** menu) | Puts the same person into a new scene | 8 per image |

These cards open other tools instead:

* **Edit Image with AI**, **Sketch to Image**, and **Style Transfer** open AI Images ready to edit. Add your photo and describe the result you want. See [Edit images](/docs/ai-images/edit).
* **Creative Upscale** opens the [Magic Upscaler](/docs/ai-images/upscale).
* **Animate Image** opens [AI Videos](/docs/videos/generate) to turn the photo into a video.
* **Headshot Generator**, **Train my AI Model**, and **Interior & Exterior Design** open [Headshots](/docs/ai-images/headshots), [AI Photo Studio](/docs/ai-images/photo-studio), and [Interior Designer](/docs/ai-images/interior-designer).

<Columns>
  <Card title="Remove or replace a background" icon="scissors" href="/docs/ai-images/remove-background">
    Step by step, with examples.
  </Card>

  <Card title="Image models and credits" icon="table" href="/docs/ai-images/models">
    Prices for every model.
  </Card>
</Columns>


# Upscale images
Source: https://easy-peasy.ai/docs/ai-images/upscale

Make an image sharper and two or four times larger, add detail, or restore an old photo with the Magic Upscaler.

The Magic Upscaler enlarges an image and sharpens it, so it holds up in print or on a big screen. It can also add fresh detail, or clean up an old, faded photo.

<Frame>
  <img alt="The Magic Upscaler with Mode set to Crisp and Scale Factor 2, a before-and-after slider over the yoga studio image, and Upscale and Download buttons" />
</Frame>

<Steps>
  <Step title="Open the Magic Upscaler">
    Choose **Images** > **Magic Upscaler** from the menu at the top, and upload a JPG, PNG, or WebP image up to 2,000 pixels wide.

    To upscale an image you made, hover over it and click the sparkles button (**Improve**), or open it and click **Upscale image**.
  </Step>

  <Step title="Choose a mode">
    | Mode | Use it for |
    | - | - |
    | **Crisp** | A sharper, larger copy that stays true to the original. Start here. |
    | **Creative** | Adding new detail and texture. Sliders for **Creativity**, **HDR**, **Resemblance**, and **Fractality** control how far it can stray from the original. |
    | **Restore old photo** | Old, blurry, or damaged photos. |
  </Step>

  <Step title="Choose the size and upscale">
    Pick a **Scale Factor** of 2 or 4, then click **Upscale**. Drag the slider over the result to compare **Before** and **After**, and click **Download** to save it.
  </Step>
</Steps>

The upscaled image is also saved in **History** > **Images**. When you upscale one of your own results, it gets a **Download upscaled image** button next to the original.

## Credits and plans

Upscaling costs 2 credits at 2× and 4 credits at 4×. The Magic Upscaler is available on paid plans.


# Chat Completions (OpenAI-compatible)
Source: https://easy-peasy.ai/docs/api-reference/endpoint/chat-completions

POST /api/chat/completions
Use the OpenAI SDK’s chat.completions.create method with the Easy-Peasy.AI base URL and API key. Supports the request fields documented here, text responses, and SSE streaming. Multimodal support depends on the selected model. Tool calling, response_format, n, and other undocumented OpenAI options are not implemented by this endpoint.

## OpenAI SDK compatibility

Use the OpenAI SDK’s `chat.completions.create` method with the `baseURL` and `apiKey` below. This endpoint implements the request fields in this reference; tool calling, `response_format`, `n`, and other undocumented OpenAI options are not implemented.

<CodeGroup>
  ```javascript Node.js theme={null}
  import OpenAI from 'openai';

  const client = new OpenAI({
    apiKey: 'YOUR_EASY_PEASY_API_KEY',
    baseURL: 'https://easy-peasy.ai/api',
  });

  // Non-streaming
  const response = await client.chat.completions.create({
    model: 'gemini-3-flash',
    messages: [
      { role: 'system', content: 'You are a helpful assistant.' },
      { role: 'user', content: 'Hello!' },
    ],
  });
  console.log(response.choices[0].message.content);

  // Streaming
  const stream = await client.chat.completions.create({
    model: 'gemini-3-flash',
    messages: [{ role: 'user', content: 'Tell me a story.' }],
    stream: true,
  });
  for await (const chunk of stream) {
    process.stdout.write(chunk.choices[0]?.delta?.content || '');
  }
  ```

  ```python Python theme={null}
  from openai import OpenAI

  client = OpenAI(
      api_key="YOUR_EASY_PEASY_API_KEY",
      base_url="https://easy-peasy.ai/api",
  )

  response = client.chat.completions.create(
      model="gemini-3-flash",
      messages=[
          {"role": "system", "content": "You are a helpful assistant."},
          {"role": "user", "content": "Hello!"},
      ],
  )
  print(response.choices[0].message.content)
  ```

  ```bash cURL theme={null}
  curl -X POST https://easy-peasy.ai/api/chat/completions \
    -H "Content-Type: application/json" \
    -H "x-api-key: YOUR_API_KEY" \
    -d '{
      "messages": [
        {"role": "user", "content": "Hello!"}
      ],
      "model": "gemini-3-flash"
    }'
  ```
</CodeGroup>

## Authentication

This endpoint supports two authentication methods:

* **x-api-key header**: `x-api-key: YOUR_API_KEY`
* **Authorization header**: `Authorization: Bearer YOUR_API_KEY` (OpenAI SDK default)

## Model IDs and aliases

The REST default is `gemini-3.8-flash`. The `gemini-3-flash` alias used in these examples follows the current Flash model and currently resolves to that default. Use an explicit version when you need to avoid a floating alias.

These IDs are routed by the chat endpoint; model availability, limits, and multimodal support depend on the provider.

| Provider | Current model IDs |
| - | - |
| Google | `gemini-3.8-flash`, `gemini-3.7-flash`, `gemini-3.6-flash`, `gemini-3.5-flash`, `gemini-3.1-pro`, `gemini-3-pro` |
| Anthropic | `claude-opus-5`, `claude-sonnet-5`, `claude-fable-5`, `claude-fable-5-1`, `claude-opus-4-8`, `claude-opus-4-7`, `claude-opus-4-6`, `claude-opus-4-5`, `claude-sonnet-4-6`, `claude-sonnet-4-5`, `claude-haiku-4-5` |
| OpenAI | `gpt-6-astra`, `gpt-6-astra-max`, `gpt-5.6-sol`, `gpt-5.6-sol-max`, `gpt-5.6-terra`, `gpt-5.6-terra-max`, `gpt-5.6-luna`, `gpt-5.6-luna-max`, `gpt-5.5-instant`, `gpt-5.5-thinking`, `gpt-5.5-pro`, `gpt-5.4-instant`, `gpt-5.4-thinking`, `gpt-5.4-pro`, `gpt-5`, `gpt-5-mini` |
| DeepSeek | `deepseek-v4-pro`, `deepseek-v4-flash` |
| Moonshot | `kimi-k3`, `kimi-k2.7-code`, `kimi-k2.6` |
| Z.ai | `glm-5p3`, `glm-5p3-max`, `glm-5p3-flash`, `glm-5p2`, `glm-5p1`, `glm-5` |
| MiniMax | `minimax-m3` |
| Meta | `muse-spark-1.3`, `muse-spark-1.3-max` |
| Qwen | `qwen3p8-max` |
| xAI | `grok-4` |

Some older names select replacements, rather than the model their name suggests:

| Legacy ID | Resolves to |
| - | - |
| `deepseek-v3`, `deepseek-r1`, `deepseek-chat`, `deepseek-reasoner` | `deepseek-v4-flash` |
| `minimax-m2`, `minimax-m2p5`, `minimax-m2p7` | `minimax-m3` |
| `kimi-k2.5`, `kimi-k2-thinking`, `kimi-k2-instruct` | `kimi-k2.6` |
| `meta-llama-3.3-70b`, `llama4-maverick-instruct-basic` | `muse-spark-1.3` |
| `qwen3p6-plus`, `qwen3p7-plus` | `qwen3p8-max` |

Choose a listed ID. Unknown names may fall back to the default or reach a provider that rejects them. Parameters such as `temperature`, `top_p`, and `stop` are also model-dependent.

## Multimodal Messages

You can send images and audio alongside text using the OpenAI multimodal message format when the selected model supports that input. Accepting the message format does not make every model multimodal. Choose an image-capable model for vision and an audio-capable model for audio input.

### Vision (Image Input)

Send images as URLs or base64 data URIs:

<CodeGroup>
  ```javascript Node.js theme={null}
  const response = await client.chat.completions.create({
    model: 'gemini-3-flash',
    messages: [
      {
        role: 'user',
        content: [
          { type: 'text', text: 'What do you see in this image?' },
          {
            type: 'image_url',
            image_url: { url: 'https://example.com/photo.jpg' },
          },
        ],
      },
    ],
  });
  ```

  ```python Python theme={null}
  response = client.chat.completions.create(
      model="gemini-3-flash",
      messages=[
          {
              "role": "user",
              "content": [
                  {"type": "text", "text": "What do you see in this image?"},
                  {
                      "type": "image_url",
                      "image_url": {"url": "https://example.com/photo.jpg"},
                  },
              ],
          }
      ],
  )
  ```

  ```bash cURL theme={null}
  curl -X POST https://easy-peasy.ai/api/chat/completions \
    -H "Content-Type: application/json" \
    -H "x-api-key: YOUR_API_KEY" \
    -d '{
      "messages": [{
        "role": "user",
        "content": [
          {"type": "text", "text": "What do you see?"},
          {"type": "image_url", "image_url": {"url": "https://example.com/photo.jpg"}}
        ]
      }]
    }'
  ```
</CodeGroup>

Base64 images are also supported:

```json theme={null}
{
  "type": "image_url",
  "image_url": {
    "url": "data:image/png;base64,iVBORw0KGgo..."
  }
}
```

### Audio Input

Send audio as base64-encoded data (mp3, wav, webm, mp4):

```json theme={null}
{
  "role": "user",
  "content": [
    { "type": "text", "text": "Transcribe this audio." },
    {
      "type": "input_audio",
      "input_audio": {
        "data": "base64-encoded-audio-data...",
        "format": "mp3"
      }
    }
  ]
}
```

## Streaming

When `stream: true`, the response uses Server-Sent Events in OpenAI chunk format:

```
data: {"id":"chatcmpl-...","object":"chat.completion.chunk","created":...,"model":"gemini-3-flash","choices":[{"index":0,"delta":{"content":"Hello"},"finish_reason":null}]}

data: {"id":"chatcmpl-...","object":"chat.completion.chunk","created":...,"model":"gemini-3-flash","choices":[{"index":0,"delta":{},"finish_reason":"stop"}]}

data: [DONE]
```


# Chat with an agent
Source: https://easy-peasy.ai/docs/api-reference/endpoint/chat-with-bot

POST /bot/{public_url}/api
Send a message to a bot and receive a response. Each bot has a unique `public_url` identifier, available from the bot's **Integrations > API** page in the [dashboard](https://easy-peasy.ai). Uses the same `x-api-key` as all other Easy-Peasy.AI endpoints.

Supports optional conversation history for multi-turn conversations and streaming responses.

Send a message to an [agent](/docs/agents/overview) you built in the dashboard and get its answer. Agents were called Bots, which is why the path starts with `/bot`.

<Note>
  The Agents API uses a **different base URL** (`https://bots.easy-peasy.ai`) but the **same API key** as all other Easy-Peasy.AI endpoints.
</Note>

## Find your agent's ID

`public_url` is your agent's ID. In the dashboard, open your agent and go to **Channels** > **Chat bubble** > **Manage** and copy **Your Agent ID**, or open **Integrations** > **REST API** to see the full endpoint with example code.

## Streaming

When `stream: true` is set, the response is returned as **server-sent events (SSE)**. Each event contains a chunk of the agent's answer, so you can show it as it's written.

## Conversation history

To keep context across messages, send the earlier messages in `history`. Each entry has a `role` (`human` or `ai`) and a `text` field. Each request is answered on its own, so without `history` the agent doesn't know what was said before.

## What the API doesn't do

The agent answers from its knowledge and instructions. [Actions](/docs/agents/actions), such as buttons, forms, and bookings, don't run through the API.


# Create Transcription
Source: https://easy-peasy.ai/docs/api-reference/endpoint/create-transcription

POST /api/transcriptions
Programmatically create [AI Transcriptions](https://easy-peasy.ai/audios) from audio files.

## Complete workflow

Transcription is asynchronous — after submitting an audio file, you need to poll for results. Here's the full workflow:

### Step 1: Submit audio for transcription

Submit a publicly downloadable audio URL with this endpoint. A paid subscription and remaining monthly transcription allowance are required. The response includes a `uuid` you'll need for the next steps.

<Note>
  Transcription typically takes **1–5 minutes** depending on audio length. Save the `uuid` from the response.
</Note>

### Step 2: Poll for results

Use [Get Transcription Result](/docs/api-reference/endpoint/get-transcription) with the same account's API key. **Poll every 15–30 seconds** while `transcription_status` is `processing`.

API-key responses return `content`, `segments`, and `transcription_status` **at the top level**, without an `audio` wrapper. `done` means complete even if no speech was detected and `content` is empty. Stop on `failed`. Older records may contain transcript text without a status.

The following Node.js 20+ example stops after ten minutes. Keep the UUID if it times out so you can check the existing job later.

```javascript theme={null}
async function waitForTranscription(uuid, apiKey) {
  const deadline = Date.now() + 10 * 60 * 1000;

  while (Date.now() < deadline) {
    const response = await fetch('https://easy-peasy.ai/api/audios', {
      method: 'POST',
      headers: {
        'Content-Type': 'application/json',
        'x-api-key': apiKey,
      },
      body: JSON.stringify({ audio_id: uuid }),
      signal: AbortSignal.timeout(Math.max(1, Math.min(30000, deadline - Date.now()))),
    });
    const data = await response.json().catch(() => null);

    if (!response.ok) {
      throw new Error(`Transcription ${uuid}: HTTP ${response.status}: ${data?.error || 'Could not retrieve transcription'}`);
    }
    if (!data || data.uuid !== uuid) {
      throw new Error(`Unexpected response for transcription ${uuid}`);
    }
    if (data.transcription_status === 'failed') {
      throw new Error(`Transcription ${uuid} failed`);
    }
    if (data.transcription_status === 'done' ||
        (!data.transcription_status && data.content)) {
      return data;
    }
    if (data.transcription_status && data.transcription_status !== 'processing') {
      throw new Error(`Unknown transcription status: ${data.transcription_status}`);
    }

    await new Promise(resolve => setTimeout(resolve, Math.max(0, Math.min(15000, deadline - Date.now()))));
  }

  throw new Error(`Timed out waiting for transcription ${uuid}. Keep this UUID and check it later.`);
}
```

### Step 3: Generate AI content (optional)

Once the transcription is complete, use [Generate Audio Content](/docs/api-reference/endpoint/generate-audio-content) to create summaries, titles, action items, and more.

```bash theme={null}
curl -X POST https://easy-peasy.ai/api/generate-audio-content \
  -H "Content-Type: application/json" \
  -H "x-api-key: YOUR_API_KEY" \
  -d '{"audio_id": "YOUR_AUDIO_UUID"}'
```

You can also generate specific fields only:

```bash theme={null}
curl -X POST https://easy-peasy.ai/api/generate-audio-content \
  -H "Content-Type: application/json" \
  -H "x-api-key: YOUR_API_KEY" \
  -d '{
    "audio_id": "YOUR_AUDIO_UUID",
    "fields": ["summary", "title", "keywords"]
  }'
```

**Content generated depends on the audio type:**

| Audio Type | Fields |
| - | - |
| **Meeting** | summary, title, description, action items, keywords, timestamped overview, topics & bullets, LinkedIn post |
| **Podcast** | summary, title, description, show notes, Twitter thread, article, newsletter, keywords, questions, LinkedIn post, timestamped overview, topics & bullets |
| **Therapy Session** | summary, title, description, progress note, SOAP note, DAP note, keywords, timestamped overview, topics & bullets, LinkedIn post |


# Generate Text
Source: https://easy-peasy.ai/docs/api-reference/endpoint/generate

POST /api/generate
Programmatically generate outputs for any of the [Templates](https://easy-peasy.ai/presets). Each template (preset) accepts different parameters — see the request body schema and examples below.



# Generate Audio Content
Source: https://easy-peasy.ai/docs/api-reference/endpoint/generate-audio-content

POST /api/generate-audio-content
Generate AI-powered content from a completed transcription. Produces summaries, titles, action items, and more depending on the audio type.

**Content generated by audio type:**
- **Meeting**: summary, title, description, action items, keywords, timestamped overview, topics & bullets, LinkedIn post
- **Podcast**: summary, title, description, show notes, Twitter thread, article, newsletter, keywords, questions, LinkedIn post, timestamped overview, topics & bullets
- **Therapy Session**: summary, title, description, progress note, SOAP note, DAP note, keywords, timestamped overview, topics & bullets, LinkedIn post

You can generate all fields at once or specify individual fields with the `fields` parameter.



# Generate Image
Source: https://easy-peasy.ai/docs/api-reference/endpoint/generate-image

POST /api/generate-image
Generate images from a prompt, or edit supplied images. model selects the text-to-image generator; editModel selects the editor for Edit Image with AI. API-key requests wait for standard generation results. Check every returned image_url. Premium models and available credits depend on the account plan.

## Generate or edit an image

* **Text-to-image** — Provide `prompt` + `model`. The `model` picks the generator (e.g. `Nano Banana 2`, `OpenAI GPT Image 2.5 Flare Medium`).
* **Editing / reference image** — Provide `prompt` + `image` (or `images`). With no explicit `action`, this selects **Edit Image with AI** and uses `editModel` to select the editor. Other actions, such as background removal, have their own behavior.

<Note>
  Send **`editModel` explicitly** when editing. With an image and no `action`, omitting `editModel` selects `OpenAI GPT Image 2 Medium`. If you explicitly set `action: "Edit Image with AI"` and omit `editModel`, the default is `Qwen Image`. To select Qwen Image explicitly, include that action too. Text-to-image defaults to `model: "Nano Banana 2"`.
</Note>

The reference image does not need to be hosted on Easy-Peasy, but it **must be downloadable server-side** — some hosts (e.g. Wikimedia, or sites that block hotlinking / non-browser requests) return **HTTP 403** to automated fetches and will fail. If in doubt, host it on your own public CDN/bucket. Write the `prompt` as an instruction about the image (e.g. "change the color to yellow", "place this product on a marble counter").

### One reference image

```bash theme={null}
curl -X POST https://easy-peasy.ai/api/generate-image \
  -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "Place this product on a marble kitchen counter with soft morning light",
    "image": "https://yourcdn.com/product.png",
    "editModel": "Nano Banana 2",
    "dimensions": "1:1"
  }'
```

### Multiple reference images

Pass `images` as a **JSON-encoded string array** (not a JSON array):

```bash theme={null}
curl -X POST https://easy-peasy.ai/api/generate-image \
  -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "Put the person from the first image into the room in the second image",
    "images": "[\"https://yourcdn.com/person.png\", \"https://yourcdn.com/room.png\"]",
    "editModel": "Nano Banana 2",
    "dimensions": "1:1"
  }'
```

## Model naming notes

* **GPT Image 2** is exposed as three quality tiers — use the exact strings `OpenAI GPT Image 2 Low`, `OpenAI GPT Image 2 Medium`, or `OpenAI GPT Image 2 High`. There is no bare `"GPT Image 2"` value.
* **GPT Image 2.5** uses exact quality names such as `OpenAI GPT Image 2.5 Flare Medium` and `OpenAI GPT Image 2.5 Sunburst High`. The schemas list all available tiers.
* **New image models** include `Nano Banana 2 Lite`, `Seedream 5.0 Pro`, `Muse Image`, `Grok Imagine 2.0`, `WAN 2.7`, and others listed below. The generation and editing lists differ.
* **Reve** is `REVE` for text-to-image (`model`), and `Reve` / `Reve Fast` for editing (`editModel`). There is no `"Reve 2.0"`.
* Model names are matched **exactly** — always send a value as listed in the schema enums below. An unrecognized `editModel` returns a `400` error; an unrecognized `model` may fall back to a default generator, so a typo can silently produce the wrong model.

## Read the result

The response is an array of image records. Modern model integrations wait for results when you authenticate with an API key; the web-only `waitForResult` flag does not turn that off. Some legacy integrations, including Midjourney, return pending records with empty `image_url` values.

Check each record's `image_url`. If it is empty, save its `id` and poll [Get Image](/docs/api-reference/endpoint/get-image) every 15–30 seconds with the same account's key. Stop after a bounded timeout and retain the ID. A polling timeout does not cancel generation or justify automatically creating another paid job.

`outputs` defaults to one, but actual output count depends on the model. Midjourney returns four records per job. Credit amounts can be fractional.

DALL-E 3 and Imagen titles are no longer in this endpoint's documented catalog. Update older integrations to a listed model instead of relying on fallback behavior.


# Generate Sound Effect
Source: https://easy-peasy.ai/docs/api-reference/endpoint/generate-sound

POST /api/generate-sound
Generate AI sound effects from a text description using ElevenLabs. Returns the audio URL immediately (synchronous).

Usage is counted against your plan's character limit.



# Generate Talking Video
Source: https://easy-peasy.ai/docs/api-reference/endpoint/generate-talking-video

POST /api/generate-talking-video
Generate a talking video by animating a face image or video with speech. Provide either text + voice to generate audio, or supply your own audio file.

**Input combinations:**
- **Image + text + voice** — Generates speech from text using the specified voice, then animates the face in the image
- **Image + audio** — Uses the provided audio to animate the face in the image
- **Video + text + voice** — Generates speech from text, then lip-syncs the video
- **Video + audio** — Lip-syncs the video with the provided audio

**Models:**
- `premium` — High-quality avatar generation (VEED Fabric). Supports 480p and 720p.
- `standard` — Faster generation (WaveSpeed InfiniteTalk). Supports 480p and 720p.

Video generation is asynchronous. Use the [Get Video](/api-reference/endpoint/get-video) endpoint to poll for results.

**Note:** Requires a paid plan. Image must be at least 512x512 pixels. Audio/video max 5 minutes.

## Workflow

1. **Submit** a talking video request using this endpoint
2. **Save** the `id` from the response
3. **Poll** the [Get Video](/docs/api-reference/endpoint/get-video) endpoint every 15–30 seconds until `status` is `completed`
4. **Download** the video from the `url` field

<Note>
  Talking video generation requires a **paid plan**. Processing typically takes **1–5 minutes** depending on audio length and resolution.
</Note>

## Input combinations

| Input | Audio Source | Description |
| - | - | - |
| `image` | `text` + `voiceID` | Generates speech, then animates the face |
| `image` | `audio` | Animates the face with provided audio |
| `video` | `text` + `voiceID` | Generates speech, then lip-syncs the video |
| `video` | `audio` | Lip-syncs the video with provided audio |

## Voice IDs

Use the [Get TTS Voices](/docs/api-reference/endpoint/get-tts-voices) endpoint to discover available voice IDs. Both ElevenLabs and OpenAI voices are supported.

## Requirements

* **Image**: minimum 512x512 pixels
* **Video**: .mp4 or .mov format, 3–300 seconds
* **Audio**: max 5 minutes


# Generate Text-to-Speech
Source: https://easy-peasy.ai/docs/api-reference/endpoint/generate-tts

POST /api/generate-text-to-speech
Convert text to speech using various AI voices. The generated audio is returned as a publicly accessible URL.

**Usage Limits:**
- Free plan: Limited words per billing period
- Paid plans: Higher limits based on subscription tier



# Generate Video
Source: https://easy-peasy.ai/docs/api-reference/endpoint/generate-video

POST /api/generate-video
Generate videos from text, first/last frames, or reference media with MiniMax H3 Max, Seedance 2.5, Gemini Omni Flash 1.1, FLUX 3, WAN 3.0, Veo, Kling, and other models. The exact model title is required. See the [video model reference](/api-reference/video-models) for supported settings.

Generation is asynchronous. Save the returned id and poll [Get Video](/api-reference/endpoint/get-video) with the same account’s API key. Paid plans require sufficient credits; free access is limited to selected variants and settings.

## Choose a model

The API supports MiniMax H3 Max, MiniMax H3, Seedance 2.5, Gemini Omni Flash 1.1, FLUX 3, WAN 3.0, Grok Imagine 1.5, and other video families. Use the [video model reference](/docs/api-reference/video-models) to choose an exact variant and its supported inputs, durations, and resolutions.

<Note>
  The REST API requires `model`. The [MCP `generate_video` tool](/docs/mcp/tools) defaults to `MiniMax H3 Max` when you omit it. REST fields use camelCase, such as `aspectRatio` and `referenceVideos`; MCP uses `aspect_ratio` and `reference_videos`.
</Note>

## Create a video

Send one generation request and save its returned `id`. Generation runs asynchronously; an accepted request does not mean the video is ready. Most videos take a few minutes, and queues can take longer.

```bash theme={null}
curl --fail-with-body -X POST https://easy-peasy.ai/api/generate-video \
  -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "A cat walks along a beach at sunset, waves in the background, cinematic lighting",
    "model": "MiniMax H3 Max",
    "duration": "5",
    "aspectRatio": "16:9",
    "resolution": "768p"
  }'
```

MiniMax H3 Max accepts 5–15 whole seconds and `480p` or `768p`. MiniMax H3 accepts 4–15 whole seconds and `768p` or `2k`. Both families generate native audio; `generateAudio: false` does not disable it. Models with an audio toggle, such as Seedance 2.5, can accept `generateAudio: false`. Providing reference audio enables audio generation.

<Warning>
  Send `duration` as a string. Veo 3.1 text variants accept `"4s"`, `"6s"`, or `"8s"`; their Image and First-Last Frame variants accept only `"8s"`. Veo 3.1 does not accept five seconds. Equivalent strings such as `"8"` and `"8s"` are normalized by the API.
</Warning>

## Animate an image

Select a model with a first-image input and send `image`. Adding an image does not automatically switch a text model to its image variant.

```bash theme={null}
curl --fail-with-body -X POST https://easy-peasy.ai/api/generate-video \
  -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "The subject slowly turns toward the camera in soft morning light",
    "model": "MiniMax H3 Max Image",
    "image": "https://yourcdn.com/first-frame.png",
    "duration": "5",
    "aspectRatio": "16:9",
    "resolution": "768p"
  }'
```

Add `"tailImage": "https://yourcdn.com/last-frame.png"` when you want a last frame and the chosen variant supports it. Check the **Input** column in the [model catalog](/docs/api-reference/video-models#model-catalog); model suffixes alone do not identify every input mode.

## Use reference media

Reference variants use images to preserve subjects or style. Some also accept videos for motion or editing and audio for voice or sound guidance.

This example uses a character image, a motion clip, and a voice clip. Replace the placeholder URLs with your own media. Keep each MiniMax H3 Max reference video and audio clip between 2 and 15 seconds, and each type's combined duration at or below 15 seconds.

```bash theme={null}
curl --fail-with-body -X POST https://easy-peasy.ai/api/generate-video \
  -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "The character walks through the market with the motion from the video and speaks in the supplied voice",
    "model": "MiniMax H3 Max Reference",
    "referenceImages": [{ "url": "https://yourcdn.com/character.png" }],
    "referenceVideos": [{ "url": "https://yourcdn.com/motion.mp4" }],
    "referenceAudios": [{ "url": "https://yourcdn.com/voice.mp3" }],
    "duration": "5",
    "aspectRatio": "16:9",
    "resolution": "768p"
  }'
```

* Use `referenceImages`, `referenceVideos`, and `referenceAudios` as arrays of `{ "url": "..." }` objects.
* `referenceVideo` is a single-URL alternative. If you supply the `referenceVideos` array, it takes precedence, including when empty.
* Do not combine reference videos with `image` or `tailImage`. Audio references require a reference image or a supported reference video; they cannot accompany a first-frame `image`.
* Limits differ by variant. MiniMax H3 Max Reference supports up to 9 images; Seedance 2.5 Reference supports up to 30. See [reference limits](/docs/api-reference/video-models#reference-limits) for counts, durations, formats, and sizes.
* `referenceImages[].type` is optional. Kling O1/O3 uses `character`, `object`, or `style` to assign reference roles; other families use general image references.

All input URLs must be downloadable server-side without browser cookies. A URL that works in your browser may still reject the provider with HTTP 403. Signed URLs must remain valid through queue delays and the provider's download. Reference media can increase credit usage.

## Poll for the result

Call [Get Video](/docs/api-reference/endpoint/get-video) with the returned `id` and the same account's API key every 15–30 seconds. Read `video.url` when `video.status` is `completed`.

The following Node.js 20+ example checks HTTP errors, handles malformed responses, and stops after ten minutes. A local timeout does not cancel the generation; keep the ID so you can check it later.

```javascript theme={null}
async function waitForVideo(videoId, apiKey) {
  const deadline = Date.now() + 10 * 60 * 1000;

  while (Date.now() < deadline) {
    const response = await fetch(
      `https://easy-peasy.ai/api/get-video?video_id=${encodeURIComponent(videoId)}`,
      {
        headers: { 'x-api-key': apiKey },
        signal: AbortSignal.timeout(Math.max(1, Math.min(30000, deadline - Date.now()))),
      }
    );
    const data = await response.json().catch(() => null);

    if (!response.ok) {
      throw new Error(`Video ${videoId}: HTTP ${response.status}: ${data?.error || 'Could not retrieve video'}`);
    }
    if (data?.video?.status === 'completed' && data.video.url) {
      return data.video.url;
    }
    if (data?.video?.status !== 'processing') {
      throw new Error(`Unexpected response for video ${videoId}`);
    }

    await new Promise(resolve => setTimeout(resolve, Math.max(0, Math.min(15000, deadline - Date.now()))));
  }

  throw new Error(`Timed out waiting for video ${videoId}. Keep this ID and check it later.`);
}
```

## Handle errors and access limits

| Response | What to do |
| - | - |
| `400` | Check the exact model name, string duration, resolution, input mode, and reference limits. |
| `401` | Check your API key and account access. |
| `403` | Check your plan and remaining credits. Free accounts have [specific model and resolution limits](/docs/api-reference/video-models#free-plan-settings). |
| `404` from Get Video | Verify the ID and account. Missing or deleted jobs, including some failed generations, return 404. |
| `500` or a network timeout | Save any ID already received. Check the existing job before deciding whether to submit again. |

Get Video currently exposes only `processing` and `completed`, so do not wait for a `failed` status. Submitting a new generation can create another billable job; a polling timeout is not a reason to recreate it automatically.


# Get Image
Source: https://easy-peasy.ai/docs/api-reference/endpoint/get-image

GET /api/get-image
Retrieve an existing image by id using the same account’s API key. Useful for legacy image integrations that return a pending record. The query parameter is imageId (camelCase). The response is a single object; read image_url at the top level.

Use this endpoint when Generate Image returns an image record whose `image_url` is empty, or to retrieve an existing image. Pass its `id` as the camelCase query parameter `imageId`.

```bash theme={null}
curl --fail-with-body 'https://easy-peasy.ai/api/get-image?imageId=12345' \
  -H 'x-api-key: YOUR_API_KEY'
```

The response is a single object. Read `image_url` at the top level. Use the same account's API key that created the image.

For pending images, poll every 15–30 seconds with a bounded timeout. Stop on HTTP errors and retain the ID on timeout. This endpoint has no normalized job status field, so an empty URL alone cannot distinguish processing from a failed generation. Do not automatically submit another paid generation when polling stops.


# Get Preset Details
Source: https://easy-peasy.ai/docs/api-reference/endpoint/get-preset

GET /api/presets/{slug}
Retrieve the full configuration for a specific preset, including all input fields and their constraints.

Use this to understand exactly which parameters to pass when calling `/api/generate` with this preset.



# Get Transcription Result
Source: https://easy-peasy.ai/docs/api-reference/endpoint/get-transcription

POST /api/audios
Retrieve a transcription by its UUID using the same account’s API key. Fields are returned at the top level (no audio wrapper). Poll every 15–30 seconds while processing. transcription_status: done means complete, including no-speech results with empty content; failed means stop. Older records can have content without a status.

Send `{ "audio_id": "YOUR_AUDIO_UUID" }` with the same account's API key used to create the transcription. The response fields are at the top level: read `content` and `transcription_status`, rather than `audio.content`.

| Status | Meaning |
| - | - |
| `processing` | Still running; check again in 15–30 seconds. |
| `done` | Complete, including recordings with no speech and empty content. |
| `failed` | Processing failed; stop polling. |
| Missing or `null` | Older record; non-empty content indicates an available transcript. |

See the [complete polling example](/docs/api-reference/endpoint/create-transcription#step-2-poll-for-results). Retain the UUID on timeout. HTTP 404 means the record is unavailable to this account.


# Get TTS Configuration
Source: https://easy-peasy.ai/docs/api-reference/endpoint/get-tts-config

GET /api/get-text-to-speech-config
Retrieve available filter options for text-to-speech voices.

Returns the available languages, accents, genders, ages and use cases, and the total number of ready-made voices.

This endpoint helps you understand what filter values you can use when calling the `/api/get-text-to-speech-voices` endpoint.

**Workflow:**
1. Call without parameters to get all available languages
2. Choose a language
3. Call with `language` parameter to get accents available for that language
4. Use language and accent to filter voices in `/api/get-text-to-speech-voices`

**Note:** This endpoint does not require authentication.



# Get TTS Voices
Source: https://easy-peasy.ai/docs/api-reference/endpoint/get-tts-voices

GET /api/get-text-to-speech-voices
Retrieve a list of available voices for text-to-speech generation.

Returns all predefined voices (ElevenLabs and OpenAI) and optionally the user's custom cloned voices.

**Filtering:**
- Filter by language using the `language` query parameter
- Filter by accent using the `accent` query parameter
- Filter ready-made voices by `gender`, `age` and `use_case` (values from `/api/get-text-to-speech-config`)
- Filters can be combined

**Authentication:**
- Optional: Endpoint works without authentication for public voices
- Required for custom voices: Include `x-api-key` header



# Get Video
Source: https://easy-peasy.ai/docs/api-reference/endpoint/get-video

GET /api/get-video
Retrieve the status and result of a video generation request. Since video generation is asynchronous (typically 1–5 minutes), poll this endpoint every 15–30 seconds until the `status` is `completed` and the `url` field contains the video URL.

Pass the `id` returned by Generate Video or Generate Talking Video as the `video_id` query parameter. Use the same account's API key.

Read `video.status` and `video.url`. The endpoint exposes `processing` and `completed`; it does not expose a `failed` status. HTTP 404 means the job is missing, deleted, or unavailable to this account, including some failed jobs.

See the [bounded polling example](/docs/api-reference/endpoint/generate-video#poll-for-the-result). A client timeout does not cancel the job.


# List Presets
Source: https://easy-peasy.ai/docs/api-reference/endpoint/list-presets

GET /api/presets
Retrieve a list of all available text generation presets (templates). Use this to discover which presets you can pass to the `/api/generate` endpoint.

Optionally filter by category. Only presets with input fields are returned.



# Introduction
Source: https://easy-peasy.ai/docs/api-reference/introduction

Getting started with the Easy-Peasy.AI API

## Overview

The Easy-Peasy.AI API lets you programmatically access our AI-powered tools:

* **Presets** — Discover available text generation templates and their input fields
* **Text Generation** — Generate blog posts, social media captions, emails, marketing copy, and more using text templates
* **Image Generation** — Generate and edit images with Nano Banana, GPT Image, FLUX, Seedream, and other models
* **Video Generation** — Generate videos with MiniMax H3 Max, Seedance 2.5, Gemini Omni, Veo, Kling, and other [video models](/docs/api-reference/video-models)
* **Transcription** — Transcribe audio files with speaker detection and enhanced quality
* **Text-to-Speech** — Convert text to natural-sounding speech with public and custom voices
* **Chat Completions** — OpenAI-compatible chat completions endpoint (works with OpenAI SDK)
* **Bots** — Chat with your custom AI bots via API

<Card title="Using an AI agent?" icon="plug" href="/docs/mcp/introduction">
  Easy-Peasy.AI also runs an [MCP server](/docs/mcp/introduction) at `https://easy-peasy.ai/mcp` — connect Claude, Cursor, Windsurf, or any MCP client without writing API code.
</Card>

## Authentication

Authenticated REST requests use your API key in the `x-api-key` header. [Chat Completions](/docs/api-reference/endpoint/chat-completions) also accepts `Authorization: Bearer YOUR_API_KEY`. TTS Config and public TTS Voices work without a key; custom voices require authentication. The separate Bots API documents its own headers.

Keep API keys on your server, outside browser bundles and public repositories.

```bash theme={null}
curl -X POST https://easy-peasy.ai/api/generate \
  -H "Content-Type: application/json" \
  -H "x-api-key: YOUR_API_KEY" \
  -d '{"preset": "custom-generator", "keywords": "Hello world"}'
```

<Card title="Get your API key" icon="key" href="https://easy-peasy.ai/settings/api">
  Generate an API key from your Easy-Peasy.AI settings page.
</Card>

## Base URL

All API requests are made to:

```
https://easy-peasy.ai
```

The **Bots API** uses a separate base URL:

```
https://bots.easy-peasy.ai
```

## Text generation workflow

1. **Discover presets** — Call `GET /api/presets` to list available templates
2. **Get field requirements** — Call `GET /api/presets/{slug}` to see what fields a preset needs
3. **Generate content** — Call `POST /api/generate` with the preset slug and required fields

## Media generation workflow

1. **Choose the exact model and inputs.** For video, consult the [model reference](/docs/api-reference/video-models). Plan access and credit requirements depend on the model and settings.
2. **Submit once and save the ID.** Videos and transcriptions run asynchronously. Standard image integrations wait for results, while some legacy integrations return pending records.
3. **Read or poll the result.** Follow each endpoint's response shape and polling example. Keep existing IDs when a local timeout occurs; submitting again can create another billable job.

## Endpoints

<Columns>
  <Card title="List Presets" icon="list" href="/docs/api-reference/endpoint/list-presets">
    Discover available text generation templates.
  </Card>

  <Card title="Get Preset Details" icon="circle-info" href="/docs/api-reference/endpoint/get-preset">
    Get input fields and constraints for a preset.
  </Card>

  <Card title="Generate Text" icon="pen" href="/docs/api-reference/endpoint/generate">
    Generate content using any preset.
  </Card>

  <Card title="Generate Image" icon="image" href="/docs/api-reference/endpoint/generate-image">
    Generate or edit images using an exact model name.
  </Card>

  <Card title="Generate Video" icon="video" href="/docs/api-reference/endpoint/generate-video">
    Generate videos from text, images, or reference media.
  </Card>

  <Card title="Create Transcription" icon="microphone" href="/docs/api-reference/endpoint/create-transcription">
    Transcribe audio files to text.
  </Card>

  <Card title="Text-to-Speech" icon="volume-high" href="/docs/api-reference/endpoint/generate-tts">
    Convert text to natural speech.
  </Card>

  <Card title="Chat Completions" icon="comments" href="/docs/api-reference/endpoint/chat-completions">
    OpenAI-compatible chat completions endpoint.
  </Card>

  <Card title="Chat with Bot" icon="robot" href="/docs/api-reference/endpoint/chat-with-bot">
    Interact with your custom AI bots via API.
  </Card>

  <Card title="MCP Server" icon="plug" href="/docs/mcp/introduction">
    Connect AI agents with the Model Context Protocol.
  </Card>
</Columns>


# Video models
Source: https://easy-peasy.ai/docs/api-reference/video-models

Exact API model names, input modes, durations, resolutions, and reference limits

Choose an exact model name for [Generate Video](/docs/api-reference/endpoint/generate-video). Every request requires a non-empty `prompt` and a `model`.

The tables below are generated from the app's video catalog. They describe the settings accepted by the API's model validation; provider availability and plan limits also apply.

## Choose settings

* Send `duration` as a **string**, such as `"5"` or `"8s"`. A range means every whole second in that range. A comma-separated list means only those durations.
* Send an explicitly listed `resolution` and `aspectRatio` when the model offers them. `2k` and `4k` use a lowercase `k`. **Omit** a setting marked “Provider-specific” or “From input”.
* If you omit `duration`, the API selects five seconds when listed, otherwise the first configured duration, or five seconds when no duration list is configured. Provider-specific models may use their own output length. Explicit settings make requests more predictable.
* “First image” requires `image`. A model that accepts a last image also accepts `tailImage` alongside `image`. Supply both when you want first-to-last-frame interpolation.
* Reference models take `referenceImages`, or `referenceVideos` where listed. The reference tables below list optional audio inputs and limits.

<Note>
  Model suffixes are not a reliable way to select the input mode. `Grok Imagine 1.5`, `Kling 2.5 Turbo Standard`, and `Seedance 1.0 Pro Fast` require an image. `Grok Imagine 1.5 Text` and `Seedance v1 Pro Fast` take text.
</Note>

## Model catalog

### Veo

| Model | Input | Duration (seconds) | Resolution | Aspect ratio |
| - | - | - | - | - |
| `Veo 3.1 Fast` | Text | 4, 6, 8 | `720p`, `1080p`, `4k` | `16:9`, `9:16` |
| `Veo 3.1 Lite` | Text | 4, 6, 8 | `720p`, `1080p` | `16:9`, `9:16` |
| `Veo 3.1 Lite Image` | First image | 8 | `720p`, `1080p` | `16:9`, `9:16` |
| `Veo 3.1 Fast Image` | First image | 8 | `720p`, `1080p`, `4k` | `16:9`, `9:16` |
| `Veo 3.1` | Text | 4, 6, 8 | `720p`, `1080p`, `4k` | `16:9`, `9:16` |
| `Veo 3.1 Image` | First image | 8 | `720p`, `1080p`, `4k` | `16:9`, `9:16` |
| `Veo 3.1 First-Last Frame` | First image + optional last image | 8 | `720p`, `1080p`, `4k` | `16:9`, `9:16` |
| `Veo 3.1 Fast First-Last Frame` | First image + optional last image | 8 | `720p`, `1080p`, `4k` | `16:9`, `9:16` |

### Gemini Omni

| Model | Input | Duration (seconds) | Resolution | Aspect ratio |
| - | - | - | - | - |
| `Gemini Omni Flash` | Text | 4–10 (whole seconds) | Provider-specific | `16:9`, `9:16` |
| `Gemini Omni Flash Reference` | Reference images or video | Provider-specific | Provider-specific | From input |
| `Gemini Omni Flash 1.1` | Text | 3–10 (whole seconds) | `360p`, `720p`, `1080p`, `4k` | `16:9`, `9:16` |
| `Gemini Omni Flash 1.1 Reference` | Reference images or video | 3–10 (whole seconds) | `360p`, `720p`, `1080p`, `4k` | From input |
| `Gemini Omni Flash Image` | First image | 4–10 (whole seconds) | Provider-specific | From input |
| `Gemini Omni Flash 1.1 Image` | First image | 3–10 (whole seconds) | `360p`, `720p`, `1080p`, `4k` | From input |
| `Gemini Omni Flash 1.1 First-Last Frame` | First image + optional last image | 3–10 (whole seconds) | `360p`, `720p`, `1080p`, `4k` | From input |

### MiniMax

| Model | Input | Duration (seconds) | Resolution | Aspect ratio |
| - | - | - | - | - |
| `MiniMax H3 Max` | Text | 5–15 (whole seconds) | `768p`, `480p` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9` |
| `MiniMax H3 Max Reference` | Reference images or video | 5–15 (whole seconds) | `768p`, `480p` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9` |
| `MiniMax H3` | Text | 4–15 (whole seconds) | `768p`, `2k` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9` |
| `MiniMax H3 Reference` | Reference images or video | 4–15 (whole seconds) | `768p`, `2k` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9` |
| `MiniMax H3 Max Image` | First image + optional last image | 5–15 (whole seconds) | `768p`, `480p` | From input |
| `MiniMax H3 Image` | First image + optional last image | 4–15 (whole seconds) | `768p`, `2k` | `16:9` |

### Grok

| Model | Input | Duration (seconds) | Resolution | Aspect ratio |
| - | - | - | - | - |
| `Grok Imagine` | Text | 5–15 (whole seconds) | `480p`, `720p` | `16:9`, `4:3`, `1:1`, `9:16`, `3:4`, `3:2`, `2:3` |
| `Grok Imagine Reference` | Reference images | 5–10 (whole seconds) | `480p`, `720p` | `16:9`, `4:3`, `1:1`, `9:16`, `3:4`, `3:2`, `2:3` |
| `Grok Imagine 1.5 Text` | Text | 5–15 (whole seconds) | `480p`, `720p`, `1080p` | `16:9`, `4:3`, `1:1`, `9:16`, `3:4`, `3:2`, `2:3` |
| `Grok Imagine 1.5 Reference` | Reference images | 5–10 (whole seconds) | `480p`, `720p` | `16:9`, `4:3`, `1:1`, `9:16`, `3:4`, `3:2`, `2:3` |
| `Grok Imagine Image` | First image | 5–15 (whole seconds) | `480p`, `720p` | From input |
| `Grok Imagine 1.5` | First image | 5–15 (whole seconds) | `480p`, `720p`, `1080p` | From input |

### Seedance

| Model | Input | Duration (seconds) | Resolution | Aspect ratio |
| - | - | - | - | - |
| `Seedance 2.0 Fast` | Text | 4–15 (whole seconds) | `480p`, `720p`, `1080p`, `4k` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9` |
| `Seedance 2.5` | Text | 4, 5, 6, 8, 10, 12, 15, 20, 25, 30 | `480p`, `720p` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9` |
| `Seedance 2.5 Turbo` | Text | 4, 5, 6, 8, 10, 12, 15, 20, 25, 30 | `720p`, `1080p` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9` |
| `Seedance 2.5 Turbo Reference` | Reference images or video | 4, 5, 6, 8, 10, 12, 15, 20, 25, 30 | `720p`, `1080p` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9` |
| `Seedance 2.5 Reference` | Reference images or video | 4, 5, 6, 8, 10, 12, 15, 20, 25, 30 | `480p`, `720p` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9` |
| `Seedance 2.0 Turbo` | Text | 4–15 (whole seconds) | `720p`, `1080p` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9` |
| `Seedance 2.0` | Text | 4–15 (whole seconds) | `480p`, `720p`, `1080p`, `4k` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9` |
| `Seedance 2.0 Mini` | Text | 4–15 (whole seconds) | `480p`, `720p`, `1080p`, `4k` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9` |
| `Seedance 2.0 Reference` | Reference images or video | 4–15 (whole seconds) | `480p`, `720p`, `1080p`, `4k` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9` |
| `Seedance 2.0 Fast Reference` | Reference images or video | 4–15 (whole seconds) | `480p`, `720p`, `1080p`, `4k` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9` |
| `Seedance 2.0 Mini Reference` | Reference images or video | 4–15 (whole seconds) | `480p`, `720p`, `1080p`, `4k` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9` |
| `Seedance 2.0 Image` | First image + optional last image | 4–15 (whole seconds) | `480p`, `720p`, `1080p`, `4k` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9` |
| `Seedance 2.0 Fast Image` | First image + optional last image | 4–15 (whole seconds) | `480p`, `720p`, `1080p`, `4k` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9` |
| `Seedance 2.5 Turbo Image` | First image + optional last image | 4, 5, 6, 8, 10, 12, 15, 20, 25, 30 | `720p`, `1080p` | `16:9` |
| `Seedance 2.5 Image` | First image + optional last image | 4, 5, 6, 8, 10, 12, 15, 20, 25, 30 | `480p`, `720p` | `16:9` |
| `Seedance 2.0 Turbo Image` | First image + optional last image | 4–15 (whole seconds) | `720p`, `1080p` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9` |
| `Seedance 2.0 Mini Image` | First image + optional last image | 4–15 (whole seconds) | `480p`, `720p`, `1080p`, `4k` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9` |
| `Seedance 1.5 Pro` | Text | 4–12 (whole seconds) | `480p`, `720p`, `1080p` | `16:9`, `4:3`, `1:1`, `3:4`, `9:16`, `21:9` |
| `Seedance 1.5 Pro Image` | First image | 4–12 (whole seconds) | `480p`, `720p`, `1080p` | `16:9`, `4:3`, `1:1`, `3:4`, `9:16`, `21:9` |
| `Seedance 1.5 Pro First-Last Frame` | First image + optional last image | 4–12 (whole seconds) | `480p`, `720p`, `1080p` | `16:9`, `4:3`, `1:1`, `3:4`, `9:16`, `21:9` |
| `Seedance v1 Pro Fast` | Text | 3–12 (whole seconds) | Provider-specific | `1:1`, `3:4`, `9:16`, `16:9`, `4:3`, `21:9` |
| `Seedance 1.0 Pro Fast` | First image | 3–12 (whole seconds) | Provider-specific | `1:1`, `3:4`, `9:16`, `16:9`, `4:3`, `21:9` |

### FLUX

| Model | Input | Duration (seconds) | Resolution | Aspect ratio |
| - | - | - | - | - |
| `FLUX 3` | Text | 5, 6, 8, 10, 12, 15, 20 | `720p`, `1080p` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9` |
| `FLUX 3 Image` | First image + optional last image | 5, 6, 8, 10, 12, 15, 20 | `720p`, `1080p` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9` |

### WAN

| Model | Input | Duration (seconds) | Resolution | Aspect ratio |
| - | - | - | - | - |
| `WAN 3.0` | Text | 5, 8, 10, 12, 15, 20, 25, 30 | `480p`, `720p`, `1080p` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4` |
| `WAN 3.0 Reference` | Reference images | 5, 8, 10, 12, 15, 20, 25, 30 | `480p`, `720p`, `1080p` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4` |
| `WAN 3.0 Image` | First image + optional last image | 5, 8, 10, 12, 15, 20, 25, 30 | `480p`, `720p`, `1080p` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4` |
| `Wan 2.5` | Text | Provider-specific | Provider-specific | `16:9`, `9:16`, `1:1` |
| `Wan v2.2 Turbo` | Text | Provider-specific | Provider-specific | `16:9`, `9:16` |
| `Wan v2.2` | Text | Provider-specific | Provider-specific | `16:9`, `9:16` |
| `Wan 2.5 Image` | First image | Provider-specific | Provider-specific | `16:9`, `9:16`, `1:1` |
| `Wan v2.2 Turbo Image` | First image + optional last image | Provider-specific | Provider-specific | `16:9`, `9:16` |
| `Wan-2.2 Image` | First image + optional last image | Provider-specific | Provider-specific | `1:1`, `16:9`, `9:16` |

### Kling

| Model | Input | Duration (seconds) | Resolution | Aspect ratio |
| - | - | - | - | - |
| `Kling 2.6 Pro` | Text | 5, 10 | Provider-specific | `16:9`, `9:16`, `1:1` |
| `Kling 2.6 Pro Image` | First image + optional last image | 5, 10 | Provider-specific | `16:9`, `9:16`, `1:1` |
| `Kling 3.0 Pro` | Text | 3–15 (whole seconds) | Provider-specific | `16:9`, `9:16`, `1:1` |
| `Kling 3.0 Pro Image` | First image + optional last image | 3–15 (whole seconds) | Provider-specific | `16:9`, `9:16`, `1:1` |
| `Kling 3.0 Standard` | Text | 3–15 (whole seconds) | Provider-specific | `16:9`, `9:16`, `1:1` |
| `Kling 3.0 Standard Image` | First image + optional last image | 3–15 (whole seconds) | Provider-specific | `16:9`, `9:16`, `1:1` |
| `Kling O3 Pro` | Text | 3–15 (whole seconds) | Provider-specific | `16:9`, `9:16`, `1:1` |
| `Kling O3 Pro Image` | First image + optional last image | 3–15 (whole seconds) | Provider-specific | `16:9`, `9:16`, `1:1` |
| `Kling O3 Standard` | Text | 3–15 (whole seconds) | Provider-specific | `16:9`, `9:16`, `1:1` |
| `Kling O3 Standard Image` | First image + optional last image | 3–15 (whole seconds) | Provider-specific | `16:9`, `9:16`, `1:1` |
| `Kling O3 Reference` | Reference images | 3–15 (whole seconds) | Provider-specific | `16:9`, `9:16`, `1:1` |
| `Kling O1 Image` | First image + optional last image | 5, 10 | Provider-specific | `1:1` |
| `Kling O1 Reference` | Reference images | 3, 5, 10 | Provider-specific | `16:9`, `9:16`, `1:1` |
| `Kling 2.5 Turbo Pro Image` | First image + optional last image | 5, 10 | Provider-specific | `16:9` |
| `Kling 2.5 Turbo Pro` | Text | 5, 10 | Provider-specific | `16:9`, `9:16`, `1:1` |
| `Kling 2.5 Turbo Standard` | First image | 5, 10 | Provider-specific | `16:9`, `9:16`, `1:1` |

### Happy Horse

| Model | Input | Duration (seconds) | Resolution | Aspect ratio |
| - | - | - | - | - |
| `Happy Horse` | Text | 3–15 (whole seconds) | `720p`, `1080p` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4` |
| `Happy Horse Image` | First image | 3–15 (whole seconds) | `720p`, `1080p` | From input |
| `Happy Horse Reference` | Reference images | 3–15 (whole seconds) | `720p`, `1080p` | `16:9`, `9:16`, `1:1`, `4:3`, `3:4` |

### LTX-2

| Model | Input | Duration (seconds) | Resolution | Aspect ratio |
| - | - | - | - | - |
| `LTX-2 Pro` | Text | 6, 8, 10 | Provider-specific | `16:9` |
| `LTX-2 Fast` | Text | 6, 8, 10, 12, 14, 16, 18, 20 | Provider-specific | `16:9` |
| `LTX-2 Pro Image` | First image | 6, 8, 10 | Provider-specific | `16:9` |
| `LTX-2 Fast Image` | First image | 6, 8, 10, 12, 14, 16, 18, 20 | Provider-specific | `16:9` |

### Hailuo

| Model | Input | Duration (seconds) | Resolution | Aspect ratio |
| - | - | - | - | - |
| `Hailuo 2.0` | Text | 6, 10 | Provider-specific | `16:9` |
| `Hailuo 2.0 Pro` | Text | Provider-specific | Provider-specific | `16:9` |
| `Hailuo 2.3` | Text | 6, 10 | Provider-specific | `16:9` |
| `Hailuo 2.3 Pro` | Text | 5 | Provider-specific | `16:9` |
| `Hailuo 2.3 Image` | First image | 6, 10 | Provider-specific | `16:9` |
| `Hailuo 2.3 Pro Image` | First image | 5 | Provider-specific | `16:9` |
| `Hailuo 2.3-Fast Pro Image` | First image | 6 | Provider-specific | `16:9` |
| `Hailuo 2.0 Image` | First image + optional last image | 6, 10 | Provider-specific | `16:9` |
| `Hailuo 2.0 Pro Image` | First image + optional last image | Provider-specific | Provider-specific | `16:9` |

### Pixverse

| Model | Input | Duration (seconds) | Resolution | Aspect ratio |
| - | - | - | - | - |
| `Pixverse v5.5` | Text | 5, 10 | Provider-specific | `16:9`, `4:3`, `1:1`, `3:4`, `9:16` |
| `Pixverse v5.5 Image` | First image | 5, 10 | Provider-specific | `16:9`, `4:3`, `1:1`, `3:4`, `9:16` |

## Reference limits

All media URLs must be downloadable without browser cookies and remain valid through any queue delay. Limits are **per request**, not per account. Additional reference media can increase the credit cost.

Do not combine `referenceVideos` with `image` or `tailImage`. Audio references require a companion reference image or a supported reference video; they cannot be used with a first-frame `image`.

| Model | Reference images (max) | Reference video | Reference audio |
| - | - | - | - |
| `Veo 3.1 Fast` | 3 | — | — |
| `Gemini Omni Flash Reference` | 5 | 1 max; 4–20s each; 20s total; 20 MiB each | — |
| `Gemini Omni Flash 1.1 Reference` | 5 | 1 max; 4–10s each; 10s total; 20 MiB each | — |
| `MiniMax H3 Max Reference` | 9 | 3 max; 2–15s each; 15s total; 20 MiB each | 3 max; 2–15s each; 15s total |
| `Grok Imagine Reference` | 7 | — | — |
| `Grok Imagine 1.5 Reference` | 7 | — | — |
| `Seedance 2.5 Turbo Reference` | 30 | 3 max; 2–30s each; 30s total; 50 MiB each | 10 max; 2–30s each; 30s total; 15 MiB each; wav/mp3 |
| `Seedance 2.5 Reference` | 30 | 3 max; 2–30s each; 30s total; 50 MiB each | 10 max; 2–30s each; 30s total; 15 MiB each; wav/mp3 |
| `WAN 3.0 Reference` | 9 | — | 3 max; 2–30s each; 30s total; 15 MiB each; wav/mp3 |
| `Seedance 2.0 Reference` | 9 | 3 max; 2–15s each; 15s total; 30 MiB each | 3 max; 15s total |
| `Seedance 2.0 Fast Reference` | 9 | 3 max; 2–15s each; 15s total; 30 MiB each | 3 max; 15s total |
| `MiniMax H3 Reference` | 9 | 3 max; 2–15s each; 15s total; 20 MiB each | 3 max; 2–15s each; 15s total; 15 MiB each; wav/mp3 |
| `Seedance 2.0 Mini Reference` | 9 | 3 max; 2–15s each; 15s total; 30 MiB each | 3 max; 15s total |
| `Veo 3.1 Fast Image` | 3 | — | — |
| `Happy Horse Reference` | 9 | — | — |
| `Kling O3 Reference` | 4 | — | — |
| `Kling O1 Reference` | 4 | — | — |
| `Veo 3.1` | 3 | — | — |
| `Veo 3.1 Image` | 3 | — | — |
| `Veo 3.1 First-Last Frame` | 3 | — | — |
| `Veo 3.1 Fast First-Last Frame` | 3 | — | — |

Veo reference images are optional on the variants listed above. For `Gemini Omni Flash Reference` video editing, the output follows the input clip's duration. Input requirements still apply when a model offers reference images alongside a first-frame mode.

## Free plan settings

Free accounts can use the variants below with sufficient remaining credits. Send the listed resolution explicitly. All other models require a paid plan.

| Model | Resolution | Maximum duration |
| - | - | - |
| `Veo 3.1 Fast` | `720p` | Model settings |
| `Seedance 2.0 Mini` | `480p` | 5 seconds |
| `Seedance 2.0 Mini Reference` | `480p` | 5 seconds |
| `Veo 3.1 Lite` | `720p` | Model settings |
| `Veo 3.1 Lite Image` | `720p` | Model settings |
| `Veo 3.1 Fast Image` | `720p` | Model settings |
| `Seedance 2.0 Mini Image` | `480p` | 5 seconds |
| `Veo 3.1 Fast First-Last Frame` | `720p` | Model settings |

## Retired and separate features

`Sora 2`, `Sora 2 Pro`, and their Image variants are no longer accepted by this endpoint. Choose a current model and update its duration, resolution, and inputs together.

Kling Motion Control uses a separate web-app workflow. Its motion-video payload is not supported by this public API endpoint.


# Audio generator
Source: https://easy-peasy.ai/docs/audio/audio-generator

Create voices, music, and soundscapes together in one clip from a single description, with optional audio or image references.

**AI Audio Generator** turns a description into finished audio: a narrator over music, two characters talking in a café, a jingle, or a whole scene with voices and sound around them. Where [text to speech](/docs/audio/text-to-speech) reads your text in one voice, the audio generator directs the whole scene.

## Create audio

<Steps>
  <Step title="Open AI Audio Generator">
    Click **Audio** > **AI Audio Generator** in the top menu, or find it in **All Tools**. The number at the top right shows how many text-to-speech characters you have left.
  </Step>

  <Step title="Describe your audio">
    Write it in **Describe your audio**, up to 2,048 characters. Set the scene, the voices and how they speak, and the music and sounds. Put spoken words in quotes.

    Not sure where to start? Pick one under **Start with an idea**, **Narration**, **Music**, **Dialogue**, or **Soundscape**, to fill in an example description and length.

    <Frame>
      <img alt="AI Audio Generator with Seed Audio 1.0 by ByteDance, a description of a calm yoga teacher saying Welcome to Example Yoga Studio over soft ambient pads and a singing bowl, Add media, Output format MP3, Target duration Auto, and Start with an idea on the right" />
    </Frame>
  </Step>

  <Step title="Add references (optional)">
    Click **Add media** to guide the result with up to 3 audio clips (MP3 or WAV, up to 30 seconds each) or 1 image (JPG, PNG, or WebP), 10 MB each. You can't mix audio and an image. Refer to the clips in your description as `@Audio1`, `@Audio2`, and `@Audio3`, for example "Use the voice from @Audio1".
  </Step>

  <Step title="Choose the format and length">
    * **Output format**: MP3, WAV, OGG Opus, or PCM.
    * **Target duration**: **Auto**, or 5 seconds to 2 minutes. It's a guide; the actual length can differ.
  </Step>

  <Step title="Click Generate audio">
    The audio is ready in about 20 seconds and appears under **Your creations**.

    <Frame>
      <img alt="Your creations with one clip from Seed Audio 1.0 as MP3, dated Sep 26, a 16-second player, Download MP3, Reuse prompt, and View prompt" />
    </Frame>
  </Step>
</Steps>

Play each creation in the list, click **Download** to save it, or **Reuse prompt** to put its description back in the form and try a variation. **View prompt** shows the full description. **Load more** shows older creations.

## Advanced settings

<Frame>
  <img alt="Advanced settings: Sample rate 24 kHz, Speed 1x, Volume 1x, and Pitch 0 semitones, with the Generate audio button and 1,000 audio characters per generation" />
</Frame>

| Setting | What it does | Default |
| - | - | - |
| **Sample rate** | From 8 to 48 kHz. Higher is better quality and a larger file | 24 kHz |
| **Speed** | From 0.5x to 2x | 1x |
| **Volume** | From 0.5x to 2x | 1x |
| **Pitch** | From 12 semitones down to 12 up | 0 |

Speed, volume, and pitch mainly shape the voices.

## What it costs

Each generation uses 1,000 text-to-speech characters, whatever its length, and only when it succeeds. You need at least 1,000 characters left; otherwise the button changes to **Upgrade to generate**. The Free plan's 100 characters aren't enough. See [Audio plans and limits](/docs/audio/limits).

You can make one clip at a time, together with the [sound effect generator](/docs/audio/sound-effects).

<Columns>
  <Card title="Text to speech" icon="microphone" href="/docs/audio/text-to-speech">
    Read a script aloud in one voice.
  </Card>

  <Card title="Sound effects" icon="volume-high" href="/docs/audio/sound-effects">
    Make a single sound effect.
  </Card>
</Columns>


# AI dubbing
Source: https://easy-peasy.ai/docs/audio/dubbing

Translate a video or audio file into another language, keeping each speaker's voice, emotion, and timing.

**AI Dubbing** translates the speech in a video or audio file and re-voices it in another language. Each speaker keeps a voice that sounds like their own, and the delivery follows the original timing. Background sound and music stay in place.

Use it to publish videos, courses, podcasts, and ads in more languages without re-recording.

## Dub a file

<Steps>
  <Step title="Open AI Dubbing">
    Click **Audio** > **AI Dubbing** in the top menu, or find it in **All Tools**. The bar at the top shows how many text-to-speech characters you've used.
  </Step>

  <Step title="Add your file or link">
    Drag a video or audio file into **Upload File or Enter URL**, up to 1 GB and 2.5 hours. Or paste a link to a video on YouTube, X, TikTok, Vimeo, and other sites.

    <Frame>
      <img alt="AI Dubbing with front-desk-call.mp3 uploaded successfully, the URL field, Project Name Front desk call, From English, To Spanish, and Advanced Settings" />
    </Frame>
  </Step>

  <Step title="Choose the languages">
    Give the project a name if you like. Set **From** to the language spoken in the file and **To** to the language you want.
  </Step>

  <Step title="Check the settings (optional)">
    Open **Advanced Settings**:

    * **Number of Speakers**: how many people speak. Leave it empty to detect them automatically.
    * **Highest resolution**: keeps the video at its best quality, for a larger file.

    **Character Usage** shows an estimate for your file.

    <Frame>
      <img alt="Advanced Settings with Number of Speakers 2 and Highest resolution off, Character Usage of about 2000 characters per minute from your Text-to-Speech quota, about 793 characters for this file, and the Create Dubbing button" />
    </Frame>
  </Step>

  <Step title="Click Create Dubbing">
    The project appears under **Recent Projects** as **Processing...**. Short files are ready in under a minute; long videos can take much longer. When it says **Ready**, click the download button to save the dubbed file.

    <Frame>
      <img alt="Recent Projects with Front desk call, English to Spanish, 794 characters, marked Ready, with a download button" />
    </Frame>
  </Step>
</Steps>

You can leave the page while a project is processing. Open **AI Dubbing** again to see its status and download it. **Recent Projects** shows your 20 most recent projects.

## Languages

Dub from and to any of these 32 languages: Arabic, Bulgarian, Chinese, Croatian, Czech, Danish, Dutch, English, Filipino, Finnish, French, German, Greek, Hindi, Hungarian, Indonesian, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Romanian, Russian, Slovak, Spanish, Swedish, Tamil, Turkish, Ukrainian, and Vietnamese.

## Tips

* **Start with clear speech.** Dubbing works best when speakers don't talk over each other and the music is quieter than the voices.
* **Set the number of speakers** when you know it. It helps keep each person's voice separate.
* **Try a short clip first.** Check the result on a minute of your video before dubbing the whole thing.

## What it costs

AI Dubbing uses your plan's text-to-speech characters: about 2,000 per minute of the original file. The estimate is shown before you start, and the final count is based on the file's length. See [Audio plans and limits](/docs/audio/limits).

<Columns>
  <Card title="Voice changer" icon="arrows-rotate" href="/docs/audio/voice-changer">
    Change the voice without translating.
  </Card>

  <Card title="Speech to text" icon="file-audio" href="/docs/audio/speech-to-text">
    Get a transcript and subtitles.
  </Card>
</Columns>


# Audio plans and limits
Source: https://easy-peasy.ai/docs/audio/limits

How many text-to-speech characters, transcriptions, and custom voices each plan includes, what each audio tool uses, when they reset, and what happens when they run out.

Text to speech and transcription have their own allowances, separate from words and image & video credits. See them on [**Settings** > **Usage**](/docs/account/usage).

| Plan | Text-to-speech characters | Transcriptions | Custom voices | Professional clones |
| - | - | - | - | - |
| Free | 100 | 2 | 1 | No |
| Starter | 25,000 | 25 | 1 | No |
| Unlimited 50 | 50,000 | 30 | 2 | Yes |
| Unlimited | 75,000 | 300 | 4 | Yes |
| Unlimited 200 | 120,000 | 600 | 6 | Yes |
| Unlimited 350 | 200,000 | 800 | 10 | Yes |
| Unlimited 500 | 300,000 | 1,000 | 16 | Yes |
| Unlimited 1000 | 700,000 | 2,000 | 20 | Yes |
| Unlimited 2000 | 1,400,000 | 4,000 | 40 | Yes |
| Unlimited 5000 | 3,500,000 | 10,000 | 100 | Yes |
| Teams | 75,000 | 1,000 | 2 | Yes |

* **Text-to-speech characters** are shared by all the audio tools:

  | Tool | Characters used |
  | - | - |
  | [Text to speech](/docs/audio/text-to-speech) | One per character of text, half with Turbo, one per 10 with **Saver** voices |
  | [Voice changer](/docs/audio/voice-changer) and [AI dubbing](/docs/audio/dubbing) | About 2,000 per minute of audio |
  | [Music generator](/docs/audio/music) | 600 per generation (two versions) |
  | [Sound effects](/docs/audio/sound-effects) and the [audio generator](/docs/audio/audio-generator) | 1,000 per generation |

  Marky's speech and music, and voiceovers and music in AI videos, use them too. The tools call them by different names on screen: **Usage**, **Speech credits**, and **audio characters**.
* **Transcriptions** are used by [speech to text](/docs/audio/speech-to-text) and [Record & Transcribe](/docs/audio/record): one per item, whatever its length.
* **Custom voices** are your [Instant clones, Professional clones, and designed voices](/docs/audio/voices) together.
* **Professional clones** shows which plans can make them.

On a Teams plan, each member has the full allowance. Compare all plans on the [pricing page](https://easy-peasy.ai/pricing).

## When they reset

Paid plans reset on your billing date each month. The Free plan resets on the 1st of the month. Unused characters and transcriptions don't carry over, and [bonus words](/docs/account/buy-credits) don't add characters or transcriptions.

## When they run out

* **Characters:** the generate button changes to an upgrade button. In text to speech and the music generator, a job that starts within your allowance finishes even if it goes over. The voice changer and AI dubbing won't start a file whose estimate is more than you have left, and sound effects and the audio generator need 1,000 characters left.
* **Transcriptions:** you can't add a new item until your allowance resets. **Upload new** and **Record Audio** change to an upgrade link, and Record & Transcribe turns recording off.
* **Custom voices:** **Clone Your Voice** and **Design a Voice** are turned off. Delete a voice you don't use to free a place.

To get more, upgrade your plan on [**Settings** > **Subscription**](/docs/account/subscription).

<Columns>
  <Card title="Plans and credits" icon="coins" href="/docs/account/plans-and-credits">
    How all the allowances work.
  </Card>

  <Card title="Usage" icon="chart-simple" href="/docs/account/usage">
    See what you've used this period.
  </Card>
</Columns>


# Music generator
Source: https://easy-peasy.ai/docs/audio/music

Make a full song or an instrumental track from a description, or from your own lyrics and style, with Suno.

**AI Music Generator** writes and produces songs: vocals, instruments, and cover art. Describe the song you want, or write your own lyrics and style. Each generation gives you two versions to choose from.

Use it for jingles, background music for videos and podcasts, birthday songs, and personal gifts.

## Make a song

<Steps>
  <Step title="Open AI Music Generator">
    Click **AI Music Generator** in the sidebar, or **Audio** > **Music Generator** in the top menu. **Usage** at the top shows how many text-to-speech characters you've used this period.
  </Step>

  <Step title="Choose a model">
    **Suno V6** is the default. You can also pick **Suno V6 Wild** or **Suno V6 Mini** and compare; all three cost the same.
  </Step>

  <Step title="Describe the song">
    With **Custom Mode** off, describe it in **Song Description**, up to 400 characters: the genre, mood, what it's about, and any names to include. The lyrics are written for you.

    For example: "An upbeat acoustic song for a yoga studio's anniversary, about morning classes, community, and breathing easy."
  </Step>

  <Step title="Set the vocals">
    Tick **Instrumental Only** for music without singing. Otherwise, set **Vocal Gender** to **Male**, **Female**, or **Auto (Let AI decide)**.
  </Step>

  <Step title="Click Generate">
    The song takes a few minutes. Two versions appear under **Music**, each with its own cover.

    <Frame>
      <img alt="The Music tab with two generated versions of the song Calm acoustic folk, soft fingerpicked guitar, each with crayon-drawing cover art and Play, Share, Details, and Download buttons" />
    </Frame>
  </Step>
</Steps>

## Write your own lyrics

Turn on **Custom Mode** to control the words and the sound:

* **Style**: genres, instruments, mood, and tempo, up to 200 characters. For example, "Calm acoustic folk, soft fingerpicked guitar, gentle female vocals, 70 BPM".
* **Lyrics**: up to 3,000 characters. Mark the parts with tags such as `[Verse]`, `[Chorus]`, and `[Bridge]`.

<Frame>
  <img alt="AI Song Generator with Model Suno V6, Custom Mode on, a Style of calm acoustic folk, Lyrics with a Verse and a Chorus about rolling out a yoga mat and breathing, Instrumental Only off, Vocal Gender Female, and Generate" />
</Frame>

On the Free plan, the description is limited to 100 characters and the lyrics to 400.

## Your songs

Each song has **Play**, **Share**, **Details** (ⓘ), and **Download** (MP3).

<Frame>
  <img alt="Song details: the cover, the text and lyrics, Created At, Characters Used 300, Status Completed, Generate Lyric Video, and Delete, Play, Share, and Download" />
</Frame>

* **Details** shows the lyrics, the date, and the characters used, and has **Delete**.
* **Generate Lyric Video** makes a simple video of the cover with the lyrics on screen. It's not a music video.
* **Share** makes a public link, like it does for [speech](/docs/audio/text-to-speech). On a Teams plan you can also share with your team.

**Examples** shows songs other people have made; open **Details** on one to see its prompt. All your songs are also listed in **History** > **Music**.

## What it costs

Each generation uses 600 text-to-speech characters: 300 for each of the two versions. If a song fails, it isn't counted. The Free plan's 100 characters are enough for one generation a month; paid plans include many more. See [Audio plans and limits](/docs/audio/limits).

<Tip>
  The **Usage** number at the top updates when you reload the page.
</Tip>

You can also ask [Marky](/docs/marky/media#music) to make a song in a chat.

<Columns>
  <Card title="Sound effects" icon="volume-high" href="/docs/audio/sound-effects">
    Make sound effects and ambience.
  </Card>

  <Card title="Audio generator" icon="waveform-lines" href="/docs/audio/audio-generator">
    Mix voices, music, and sound in one clip.
  </Card>
</Columns>


# Record & Transcribe
Source: https://easy-peasy.ai/docs/audio/record

Record a meeting, call, or lecture in your browser for up to two hours and get a transcript, summary, and notes.

**Record & Transcribe** records through your microphone and transcribes the recording when you stop. Use it for in-person meetings, lectures, coaching sessions, or calls on speaker.

<Frame>
  <img alt="Record & Transcribe: What type of recording is this? set to Meeting, Select language set to English, a Start recording button, and View past recordings" />
</Frame>

## Record

<Steps>
  <Step title="Choose the type and language">
    Click **Record & Transcribe** in the sidebar. Choose **What type of recording is this?**, such as Meeting or Lecture, and **Select language**. The type decides which [AI Content](/docs/audio/transcripts#ai-content) you get.
  </Step>

  <Step title="Click Start recording">
    Allow microphone access when the browser asks. Recording starts right away.
  </Step>

  <Step title="Pause or stop">
    Pause and resume as often as you need; paused time doesn't count. Stop when you're done. The recording uploads, and the transcript page opens.
  </Step>
</Steps>

Recordings can be up to 2 hours long; recording stops on its own at the limit. Speakers are detected automatically.

## Good to know

* **The microphone only.** It records what your microphone hears, not the sound from other apps or browser tabs. For an online meeting, play the other speakers out loud, or record in the meeting app and [upload the file](/docs/audio/speech-to-text).
* **Silence warning.** If no sound reaches the microphone for about 15 seconds, you'll see a warning. Check the input device and that the microphone isn't muted.
* **Saved as you go.** The recording uploads in 30-second pieces while you record. If the tab closes, what was recorded so far is kept; open the item from **AI Speech to Text** and click **Retry transcription** after about 15 minutes, or wait for it to be transcribed automatically.
* **Upload problems.** If pieces fail to upload when you stop, you can retry them or continue with what arrived.

## What it costs

Each recording uses one transcription, counted when you click **Start recording**. When you've used your plan's transcriptions, recording is turned off until your allowance resets or you upgrade. See [Audio plans and limits](/docs/audio/limits).

<Columns>
  <Card title="Work with a transcript" icon="file-lines" href="/docs/audio/transcripts">
    Speakers, summaries, notes, and AI Chat.
  </Card>

  <Card title="Speech to text" icon="file-audio" href="/docs/audio/speech-to-text">
    Transcribe a file or link instead.
  </Card>
</Columns>


# Sound effects
Source: https://easy-peasy.ai/docs/audio/sound-effects

Describe a sound and get a sound effect or ambience for your videos, games, apps, and podcasts, as WAV or MP3.

**Sound Effect Generator** makes sound effects and background ambience from a description: footsteps on gravel, rain on a window, a door creaking open, a sci-fi engine hum. It makes sounds only, never speech or music; for a voiceover or a song, use [text to speech](/docs/audio/text-to-speech) or the [music generator](/docs/audio/music).

## Generate a sound effect

<Steps>
  <Step title="Open Sound Effect Generator">
    Click **Audio** > **Sound Effect Generator** in the top menu, or find it in **All Tools**. The number at the top right shows how many text-to-speech characters you have left.
  </Step>

  <Step title="Describe the sound">
    Type it in **Describe the sound effect you want**, up to 2,000 characters. Say what makes the sound, where it happens, what it sounds like, and how it changes over time. **Ambience**, **Foley**, and **Sci-fi** fill in an example to start from.

    <Frame>
      <img alt="AI Sound Effect Generator with the description A Tibetan singing bowl struck once in a quiet room, the warm tone ringing out and slowly fading to silence, the Ambience, Foley, and Sci-fi examples, Target duration 8 sec, Download format WAV for editing, and Generate sound" />
    </Frame>
  </Step>

  <Step title="Set the length and format">
    * **Target duration**: 1 to 30 seconds. The AI aims for it; the actual length can differ a little.
    * **Download format**: **WAV** for editing, or **MP3** for a smaller file.
  </Step>

  <Step title="Click Generate sound">
    The sound is ready in about 20 seconds and appears under **Generated Sound Effects**.

    <Frame>
      <img alt="Generated Sound Effects with one sound, A Tibetan singing bowl struck once in a quiet room, with Play, Share, Details, and Download buttons, and a History link" />
    </Frame>
  </Step>
</Steps>

## Your sound effects

Each sound has **Play**, **Share**, **Details**, and **Download**. **Details** shows the description, the date, and the characters used, and has **Delete**. **Share** works like it does for [speech](/docs/audio/text-to-speech): **Share Publicly** makes a public link.

The page shows your 18 most recent sounds. **History** lists the rest.

## Tips for descriptions

* **Name the source and the surface.** "Boots walking slowly on wet gravel" works better than "footsteps".
* **Describe the space.** A small room, a cathedral, an open field, underwater.
* **Describe changes over time.** "Rain starts softly, gets heavier, then a distant thunder roll."
* **Keep one scene per sound.** Generate separate sounds and layer them in your editor.

## What it costs

Each sound uses 1,000 text-to-speech characters, whatever its length. Characters are used only when a sound is saved, so a failed generation costs nothing. You need at least 1,000 characters left to generate; otherwise the button changes to **Upgrade plan**. The Free plan's 100 characters aren't enough for a sound effect. See [Audio plans and limits](/docs/audio/limits).

You can make one sound at a time, together with the [audio generator](/docs/audio/audio-generator).

## For developers

Generate sound effects from your own code with `POST /api/generate-sound`. See [Generate sound effect](/docs/api-reference/endpoint/generate-sound).

<Columns>
  <Card title="Audio generator" icon="waveform-lines" href="/docs/audio/audio-generator">
    Mix voices, music, and sound in one clip.
  </Card>

  <Card title="Music generator" icon="music" href="/docs/audio/music">
    Make a song or background track.
  </Card>
</Columns>


# Speech to text
Source: https://easy-peasy.ai/docs/audio/speech-to-text

Transcribe audio and video files, YouTube, Instagram, TikTok, and Google Drive links, with speaker labels, timestamps, and subtitles.

**AI Speech to Text** turns recordings into text: meetings, interviews, podcasts, lectures, calls, and videos. The transcript shows who said what and when, and you can turn it into summaries, notes, and posts with [AI Content](/docs/audio/transcripts#ai-content).

## Transcribe a file or link

<Steps>
  <Step title="Choose a source">
    Click **AI Speech to Text** in the sidebar, then **Upload new**.

    <Frame>
      <img alt="Source options: Transcribe a YouTube Video, an Instagram Video, a TikTok Video, a Google Drive File, Import URL, Transcribe an Uploaded File, Record Audio, and Integrate with Zapier" />
    </Frame>

    | Source | What to give |
    | - | - |
    | **Transcribe an Uploaded File** | An audio or video file from your computer |
    | **Transcribe a YouTube Video** | A YouTube link, including Shorts |
    | **Transcribe an Instagram Video**, **a TikTok Video** | The post's link |
    | **Transcribe a Google Drive File** | A Google Drive link that anyone with the link can open |
    | **Import URL** | A public link to an audio or video file |
    | **Record Audio** | Opens [Record & Transcribe](/docs/audio/record) |
    | **Integrate with Zapier** | Sends recordings from other apps through Zapier |
  </Step>

  <Step title="Add the file or link">
    For a file, drag it in or click to choose it. Audio: MP3, M4A, WAV, FLAC, OGG, AAC, Opus, WMA. Video: MP4, MOV, WebM, AVI, MKV, FLV.

    <Frame>
      <img alt="The upload form with front-desk-call.mp3 uploaded, Type of Audio Phone Call, Language spoken in the audio English, and Detect Speakers on" />
    </Frame>
  </Step>

  <Step title="Set the options">
    <Frame>
      <img alt="Options: Detect Speakers on, Speakers expected 2, Enhanced Quality Transcription, Redact Personal Identifiable Information (PII), Specialized Terminology, Use Transcription 2.0 on, and a Transcribe button" />
    </Frame>

    | Option | What it does |
    | - | - |
    | **Type of Audio** | Podcast, Meeting, Lecture, Sales Call, Therapy Session, and so on. It decides which [AI Content](/docs/audio/transcripts#ai-content) you get |
    | **Language spoken in the audio** | The language, or **Auto** to detect it. Choosing the language gives better results, especially with accents |
    | **Detect Speakers?** | Labels who is speaking |
    | **Speakers expected** | How many people speak. Leave it empty to detect automatically |
    | **Redact Personal Identifiable Information (PII)** | Replaces personal details with `####`: names, organizations, email addresses, dates of birth, passport, social security, health care, account, and credit card numbers, and banking details |
    | **Use Transcription 2.0** | The current engine, on by default. It adds subtitle downloads and speaker counts. Leave it on |

    **Specialized Terminology** and **Enhanced Quality Transcription** currently only affect the older engine, used when **Use Transcription 2.0** is off.
  </Step>

  <Step title="Click Transcribe">
    The file uploads and transcription starts. Short recordings are done in a minute or two; long ones can take 10 to 20 minutes. You can leave the page; you'll get an email when it's ready.

    <Frame>
      <img alt="The transcript page for front-desk-call.mp3 with the Transcription, AI Content, and AI Chat tabs and the message Transcription is in progress" />
    </Frame>
  </Step>
</Steps>

Then [read and work with the transcript](/docs/audio/transcripts).

### YouTube videos

When a YouTube video has captions, the transcript is ready in seconds as plain text, without speakers or timestamps, and subtitle downloads aren't available. Videos without captions are transcribed like a file, with speakers and timestamps.

## Your transcriptions

**AI Speech to Text** lists your transcriptions, 20 per page, each with its status: **Processing**, **Transcribed**, **No speech**, or **Failed**. Click one, or **Edit**, to open it.

<Frame>
  <img alt="The AI Transcription page with Record Audio and Upload new buttons, and one item, Front desk call, marked Transcribed" />
</Frame>

On a Teams plan, transcriptions teammates share with the team appear here too.

## If a transcription fails

* **Failed** items are retried automatically. You can also open one and click **Retry transcription**; your file is kept.
* **No speech** means no speech was found, for example a recording with the microphone muted. Check the recording and upload it again.

## Dictate into a text box

The microphone button in Marky, the tools, and AI Images types what you say into the text box. Dictation doesn't use transcriptions.

## What it costs

Each new item uses one transcription from your plan, whatever its length. An item counts when you pick a source or start a recording, even if you don't finish it or delete it later. Retries don't count again. Check your usage on [**Settings** > **Usage**](/docs/account/usage) and see [Audio plans and limits](/docs/audio/limits).

AI Content and AI Chat use words.

## For developers

Transcribe from your own code with `POST /api/transcriptions`, which needs a paid plan. See [Create transcription](/docs/api-reference/endpoint/create-transcription).

<Columns>
  <Card title="Work with a transcript" icon="file-lines" href="/docs/audio/transcripts">
    Speakers, subtitles, AI Content, and AI Chat.
  </Card>

  <Card title="Record & Transcribe" icon="microphone" href="/docs/audio/record">
    Record a meeting in your browser.
  </Card>
</Columns>


# Text to speech
Source: https://easy-peasy.ai/docs/audio/text-to-speech

Turn text into natural speech in more than 30 languages, with expressive audio tags, AI help with the script, and downloads in MP3, WAV, or Opus.

**AI Text to Speech** turns your text into a voiceover. Pick one of about 2,000 voices, or [your own](/docs/audio/voices), paste your script, and download the audio. Use it for videos, podcasts, ads, e-learning, and phone messages.

<Frame>
  <img alt="AI Text to Speech Generator: a text box with a welcome message for Example Yoga Studio, AI writing buttons, Use Turbo, the voice Ivan - Professional with Change voice, Clone Your Voice, Design a Voice, Advanced Settings, and Generate" />
</Frame>

## Generate speech

<Steps>
  <Step title="Open AI Text to Speech">
    Click **AI Text to Speech** in the sidebar. **Usage** at the top right shows how many text-to-speech characters you've used this period.
  </Step>

  <Step title="Enter your text">
    Type or paste it in **What text do you want to convert to speech?** The counter shows how many characters you can use in one go (see [Length limits](#length-limits)).

    To get help with the text, use the buttons under the box: **Write a script** writes one from a topic, **Improve**, **Fix grammar**, and **Shorten** rewrite what you have. **Undo** brings back your previous text.
  </Step>

  <Step title="Pick a voice">
    Click **Change voice** and choose one. See [Voices and voice cloning](/docs/audio/voices).
  </Step>

  <Step title="Adjust the settings (optional)">
    Turn on **Use Turbo** to use half the characters, or open **Advanced Settings** to change the file format, speed, and delivery. See below.
  </Step>

  <Step title="Click Generate">
    The audio appears at the top of **Generated Audios**, usually within a minute. Texts over 10,000 characters are generated in the background: the card shows **Processing** and turns into a player when the audio is ready, which can take a few minutes. Click **Play** to listen.
  </Step>
</Steps>

## Audio tags

Voices marked **Audio tags** in the voice list, [designed voices](/docs/audio/voices#design-a-voice), and your own [instant clones](/docs/audio/voices#instant-clone) can act out stage directions written in square brackets, such as `[laughs]`, `[whispers]`, `[sighs]`, or `[exhales]`. With one of these voices selected, **Add audio tags** adds suitable tags to your text for you.

<Frame>
  <img alt="The text box with an [exhales] tag added after the first sentence, the tooltip Insert expressive tags like [laughs] or [whispers] where they fit, and the voice Hope selected" />
</Frame>

Other voices can't act out tags, so text in square brackets is skipped: it isn't read aloud and its characters aren't counted. If your text has tags and the voice doesn't support them, a note says so and offers **Use a voice with audio tags**. Remove the brackets from any words you do want read. At a **Speed** other than 1.00x, only designed voices and voices marked **Audio tags** (with **Use Turbo** off) act tags out, and you're offered **Set speed to 1.0x**. The default voice, Hope, supports audio tags.

## Turbo

**Use Turbo** uses a faster model and counts half the characters. Try it with your voice and compare. Turbo isn't available with [designed voices](/docs/audio/voices#design-a-voice) and [Professional clones](/docs/audio/voices#professional-clone).

## Advanced Settings

<Frame>
  <img alt="Advanced Settings: Format MP3 128kbps (Recommended), Speed 1.00x, Stability 70%, and Clarity 75%" />
</Frame>

| Setting | What it does | Default |
| - | - | - |
| **Format** | The audio file: MP3 at 64 to 192 kbps, WAV at 24 or 44.1 kHz, or Opus at 64 or 128 kbps | MP3 128kbps |
| **Speed** | From 0.7x to 1.2x. Speeds other than 1.00x use an earlier voice model, which acts out audio tags only on voices marked **Audio tags** | 1.00x |
| **Stability** | Lower (**Expressive**) varies the delivery more; higher (**Stable**) keeps it even | 70% |
| **Clarity** | How closely the audio sticks to the original voice, from **Natural** to **Maximum** | 75% |

**Reset to defaults** restores all four. Voices marked **Saver** always produce MP3 and ignore these settings and Turbo.

## Length limits

| Voice | Characters per generation |
| - | - |
| Most voices | 50,000 (20,000 in WAV format) |
| Professional clones | 10,000 |
| Voices marked **Saver** | 4,096 |
| Free plan, any voice | 100 |

Texts over 10,000 characters are generated in the background. The whole text has to fit in the text-to-speech characters left on your plan. For even longer scripts, split the text and generate it in parts.

## Your generated audio

**Generated Audios** shows your 18 newest files. All of them are in **History** > **AI Text-to-Speech**.

<Frame>
  <img alt="Generated Audios with four cards, each with the title, the voice Hope, Play, share, details, and download icons, and a player at the bottom playing the first one" />
</Frame>

Each card has **Play**, **Share**, details (ⓘ), and **Download**. Downloads are named after the voice and the date.

Details show the text, the voice, the date, and **Characters Used**, and have these buttons:

<Frame>
  <img alt="Details of a generated audio: the text, the voice Hope, Created At, Characters Used 152, Status Completed, and Delete, Regenerate, Play, Share, and Download" />
</Frame>

* **Regenerate** puts the text and voice back in the form, so you can change them and generate again. It doesn't copy the Advanced Settings.
* **Share** turns on a public page for the audio, with buttons to post it on LinkedIn, Facebook, X, and Reddit. On a Teams plan you can also share it with your team.
* **Delete** removes the audio. Deleted audio can't be restored.

## What it costs

Text to speech uses your plan's text-to-speech characters:

| Voice or option | Characters used |
| - | - |
| Most voices | One per character of text, including spaces and audio tags. Tags the voice skips aren't counted |
| **Use Turbo** | Half |
| Voices marked **Saver** | One for every 10 characters |

Characters are only counted for audio that was created; a failed generation costs nothing. The AI writing buttons (**Write a script**, **Improve**, and so on) use words, not characters. See [Audio plans and limits](/docs/audio/limits) for how many characters each plan includes and what happens when they run out.

<Note>
  On paid plans, you may use the audio commercially, even after your subscription ends. Audio made on the Free plan is for non-commercial use.
</Note>

## For developers

The API makes speech with `POST /api/generate-text-to-speech`. See [Text to speech in the API reference](/docs/api-reference/endpoint/generate-tts).

<Columns>
  <Card title="Voices and voice cloning" icon="microphone" href="/docs/audio/voices">
    Find a voice, clone yours, or design a new one.
  </Card>

  <Card title="Talking videos" icon="comment" href="/docs/videos/talking-videos">
    Put your voiceover on a presenter.
  </Card>
</Columns>


# Work with a transcript
Source: https://easy-peasy.ai/docs/audio/transcripts

Name the speakers, play the audio from any line, download SRT and VTT subtitles, generate summaries and notes with AI Content, and ask questions with AI Chat.

Open a transcription from **AI Speech to Text**. The page has three tabs: **Transcription**, **AI Content**, and **AI Chat**, and a player at the bottom.

<Frame>
  <img alt="The Front desk call transcript page: the title with a rename pencil, the Transcription, AI Content, and AI Chat tabs, a banner to generate an AI summary, and Speakers A and B named Heather and Daniel" />
</Frame>

## The transcript

<Frame>
  <img alt="Speakers Heather and Daniel, and the transcription with each line labeled by speaker with start and end times, a Download menu, a Copy button, and Delete transcription" />
</Frame>

* **Speakers**: replace Speaker A, Speaker B, and so on with real names. The names appear in the transcript and help AI Content write accurate summaries.
* **Play from a line**: click the speaker icon next to a name to play from there. The words are highlighted as they're spoken.
* **Copy** copies the whole transcript with speaker names.
* **Download** saves subtitles in **SRT Format** or **VTT Format**, with timings, for video editors and YouTube.

<Frame>
  <img alt="The Download menu with SRT Format and VTT Format" />
</Frame>

When speaker detection is off, the tab is called **Plain Text** and shows the text without speakers or times.

The transcript text can't be edited. To fix names and terms in what you publish, edit the AI Content or copy the text into your editor.

The player at the bottom plays the recording, skips 10 seconds back or forward, changes the playback speed, and downloads the recording.

<Frame>
  <img alt="The player bar with play, 10-second back and forward buttons, a progress bar at 00 of 23 seconds, and speed, volume, and download buttons" />
</Frame>

## AI Content

**AI Content** turns the transcript into ready-to-use text. The first time you open it, a summary and the other items for your type of recording are written automatically.

<Frame>
  <img alt="AI Content with a Phone Call Summary: an overview of Heather's call with Daniel, a first-time student who booked the free Sunrise Flow class, followed by Key Points" />
</Frame>

| Type of recording | What you get |
| - | - |
| Every type except Therapy Session | Summary, Title, Timestamped Overview, Topics and Bullets, Description, LinkedIn Post, Keywords |
| Podcast | Also show notes, a Twitter thread, a newsletter, an article, and more |
| Meeting | Also Action Items |
| Therapy Session | Summary, Progress Note, SOAP Note, DAP Note, Topics and Bullets, Keywords |

<Frame>
  <img alt="The AI Content items collapsed: Summary, Title, Timestamped Overview, Topics and Bullets, Description, and LinkedIn Post" />
</Frame>

Each item has **Generate** to write it again and **Copy**. **Collapse All** and **Expand All** fold the list. AI Content uses words.

## AI Chat

**AI Chat** answers questions about the recording: what was decided, who said what, or a draft of a follow-up email.

<Frame>
  <img alt="AI Chat: the question Which class did Daniel book, and does he need to bring anything? and the answer that he booked the Sunrise Flow class on Monday at 6:30 and doesn't need to bring anything" />
</Frame>

AI Chat reads about the first 10,000 tokens of the transcript, roughly 45 minutes of conversation. For longer recordings, ask about the beginning, or use AI Content, which reads more. AI Chat uses words.

## Rename, share, and delete

* **Rename**: click the pencil next to the title, type, and press Enter.
* **Share**: click **Share** at the top right. **Share Publicly** creates a public page with the transcript and AI Content. On a Teams plan, **Share with Team Members** shows it in your teammates' lists.
* **Delete**: click **Delete transcription** at the bottom and confirm. The recording is deleted and can't be restored. Deleting doesn't give the transcription back to your allowance. Turn off **Share Publicly** before you delete.

<Columns>
  <Card title="Speech to text" icon="file-audio" href="/docs/audio/speech-to-text">
    Transcribe another file or link.
  </Card>

  <Card title="Record & Transcribe" icon="microphone" href="/docs/audio/record">
    Record a meeting in your browser.
  </Card>
</Columns>


# Voice changer
Source: https://easy-peasy.ai/docs/audio/voice-changer

Turn a recording or video into a different voice while keeping your words, emotion, and timing.

**Voice Changer** re-voices a recording. You speak or act the lines; the voice changes, and your words, pacing, and emotion stay exactly as performed. It works with audio and with video: for a video, you get the video back with the new voice.

Use it to give a narration a different voice, keep one voice across takes recorded by different people, or voice a character.

## Change a voice

<Steps>
  <Step title="Open Voice Changer">
    Click **Audio** > **Voice Changer** in the top menu, or find it in **All Tools**.
  </Step>

  <Step title="Upload your audio or video">
    Drag the file into **Upload your audio or video**, or click to choose it. MP3, WAV, M4A, AAC, OGG, FLAC, MP4, MOV, and WebM files up to 100 MB and 30 minutes long work.
  </Step>

  <Step title="Choose a target voice">
    Click **Change** and pick a voice. Play a few to compare. You can use any voice from the library, and your [Instant clones and designed voices](/docs/audio/voices).

    <Frame>
      <img alt="Voice Changer with class-welcome.mp3 uploaded, 0:07 long, the target voice Matthew - Smooth, English, American, Remove background noise on, about 250 credits for this audio with 74,167 remaining, and a Transform voice button" />
    </Frame>
  </Step>

  <Step title="Click Transform voice">
    **Remove background noise**, on by default, cleans up hiss and room sound first. Turn it off to keep the original atmosphere.

    The result appears under **Your conversions** a few seconds later.

    <Frame>
      <img alt="Your conversions with one result in the voice Matthew - Smooth, dated Sep 26, 2026, 8 seconds long, with a player and a download button" />
    </Frame>
  </Step>
</Steps>

Play the result in the list, or click the download button to save it: an MP3 for audio, an MP4 for video. **Your conversions** shows your 20 most recent results; they can't be deleted or shared from this page.

## Tips

* **Record clean audio.** One speaker, close to the microphone, no music. The new voice copies your delivery, so perform the lines the way you want them to sound.
* **Match the voice to the performance.** A calm voice reading shouted lines sounds odd. Pick a voice that suits the energy of the recording.
* **Professional clones don't work here.** Voice Changer can't use [Professional clones](/docs/audio/voices#professional-clone). To change a recording into your own voice, make an [Instant clone](/docs/audio/voices#instant-clone) from the same sample.

## What it costs

Voice Changer uses your plan's text-to-speech characters, shown on this page as **Speech credits**: about 2,000 per minute of audio, or 33 per second. Before you start, the page shows an estimate for your file; the final count is based on the length of the result.

A failed conversion isn't counted. If the estimate is more than you have left, **Transform voice** changes to **Upgrade to continue**. See [Audio plans and limits](/docs/audio/limits).

<Columns>
  <Card title="Voices and voice cloning" icon="microphone" href="/docs/audio/voices">
    Clone your voice or design a new one.
  </Card>

  <Card title="AI dubbing" icon="language" href="/docs/audio/dubbing">
    Translate a video into another language.
  </Card>
</Columns>


# Voices and voice cloning
Source: https://easy-peasy.ai/docs/audio/voices

Find a voice by language, accent, gender, age, or style, clone your own voice from a recording, or design a new voice from a description.

Every text-to-speech voice comes from the voice list: about 2,000 ready-made voices, plus the voices you add yourself. Your own voices work across Easy-Peasy.AI, for example in [talking videos](/docs/videos/talking-videos) and the [Video Agent](/docs/videos/video-agent).

## Choose a voice

On [AI Text to Speech](/docs/audio/text-to-speech), click **Change voice**.

<Frame>
  <img alt="Choose a voice for the audio: a search box, Language and Accent filters, an Audio tags button, and voice cards such as Hope, Heather Rey, Matthew - Smooth, and Pro Narrator, each with a language badge, a use such as Conversational or Narrative Story, a play button, and a star" />
</Frame>

* The search box finds voices by name, description, language, accent, gender, age, or style. Every word has to match, so `calm british female` finds calm British women's voices.
* **All languages** and **All accents** narrow the list, for example English with a Singaporean accent.
* **Any gender**, **Any age**, **Any use case**, and **Any tone** filter the ready-made voices, for example young female voices for social media, or calm voices for narration. Each option shows how many voices it leaves; **Clear filters** resets everything. On a phone, tap **Filters** to show them.
* **Audio tags** shows only voices that can act out [audio tags](/docs/audio/text-to-speech#audio-tags).
* The play button plays a sample. The star marks a favourite; favourites come first in this browser, and **Favorites** shows only them.

Click a voice to choose it, then click **Select Voice**. Your own voices appear first in the list.

Two badges to know:

* **Audio tags**: the voice acts out tags like `[laughs]`.
* **Saver**: uses one character for every 10 you convert. It always makes MP3 and ignores Turbo and Advanced Settings.

## Your own voices

Add voices three ways:

| Option | How it works | Plans |
| - | - | - |
| **Instant** clone | Copies a voice from a short recording in seconds | All |
| **Professional** clone | Captures more of the speaker's character from a clean recording | Unlimited 50 and higher, Teams, Enterprise |
| **Design a Voice** | Creates a new voice from a written description | All |

All three count toward your plan's custom voices, from 1 on the Free plan. See [Audio plans and limits](/docs/audio/limits). Creating a voice doesn't use characters.

<Warning>
  Only clone your own voice, or a voice you have permission to use.
</Warning>

## Instant clone

<Steps>
  <Step title="Open Clone Your Voice">
    On AI Text to Speech, click **Clone Your Voice**. **Instant** is selected.

    <Frame>
      <img alt="Clone Your Voice with Instant selected: guidance to use one speaker, over 1 minute, with no background noise; Upload Sample Recording; and Or Record New Audio with a microphone button" />
    </Frame>
  </Step>

  <Step title="Add a recording">
    Upload an audio or video file of up to 10 MB, or click the microphone and record up to 3 minutes. Use one speaker, more than a minute of speech, and no background noise or music.
  </Step>

  <Step title="Name it and save">
    Enter a **Voice Name** and click **Add Voice**. The voice is ready to use right away.
  </Step>
</Steps>

## Professional clone

A Professional clone sounds closer to the speaker, especially their rhythm and expression.

<Frame>
  <img alt="Clone Your Voice with Professional selected: use a clean recording of one speaker, between 10 seconds and 5 minutes, with no background noise; upload MP3, M4A or WAV up to 20 MB, or record" />
</Frame>

1. Click **Clone Your Voice** and choose **Professional**.
2. Upload an MP3, M4A, or WAV file, or record. The recording must be 10 seconds to 5 minutes long. If it's longer, drag the highlighted window to choose which part to use.
3. Optionally type **What does the recording say?**: the recording is checked against it, which catches silent or unclear samples before the clone is made.
4. Name the voice and save it.

Professional clones:

* Can't be deleted by you. Contact support to remove one.
* Can't be used in the Voice Changer. Make an Instant clone of the same recording for that.
* Ignore **Stability** and **Clarity**, and don't offer Turbo.
* Are removed if your subscription ends.

## Design a voice

Describe the voice you want, and you get three versions to choose from.

<Steps>
  <Step title="Describe the voice">
    Click **Design a Voice** and write a **Description** of 10 to 1,000 characters: age, gender, accent, tone, pace, and the job it's for. **Try an example** fills in a sample, and **Best practices** has tips.

    <Frame>
      <img alt="Design a Voice with a description of a calm, warm female yoga teacher, example buttons such as Audiobook Narrator, Explainer Video, Meditation Guide, and Movie Trailer, and a Generate voice button" />
    </Frame>
  </Step>

  <Step title="Click Generate voice">
    After up to 30 seconds, you get three versions. Play each one.
  </Step>

  <Step title="Pick one and save it">
    Select your favourite, enter a **Voice Name**, and click **Save voice**.

    <Frame>
      <img alt="Pick your favorite with Voice 1, Voice 2, and Voice 3, Voice 1 selected, a Voice Name Studio Teacher, and Regenerate and Save voice buttons" />
    </Frame>
  </Step>
</Steps>

Designed voices support audio tags. They appear at the top of the voice list:

<Frame>
  <img alt="The voice list with the designed voice Studio Teacher first, marked Audio tags, followed by Hope and Heather Rey" />
</Frame>

## Delete a voice

In the voice list, click the trash icon on one of your voices. Deleting frees a place for a new custom voice. Professional clones can't be deleted this way.

<Columns>
  <Card title="Text to speech" icon="waveform-lines" href="/docs/audio/text-to-speech">
    Make audio with your voice.
  </Card>

  <Card title="Audio plans and limits" icon="gauge" href="/docs/audio/limits">
    Custom voices and characters per plan.
  </Card>
</Columns>


# Create a brand voice
Source: https://easy-peasy.ai/docs/brand-voice/create

Turn your website, writing samples, or a description of your style into writing instructions that Marky and the tools follow.

A brand voice is a set of writing instructions for your business: its tone, vocabulary, sentence style, and the words it avoids. You give Easy-Peasy.AI your website or a few samples of your writing, and it writes the instructions for you. Then [Marky and the tools](/docs/brand-voice/use) can write in that voice.

Creating a brand voice is free. You can have as many as you like, for example one for your blog and one for social media.

## Create a brand voice

<Steps>
  <Step title="Open Brand Voice">
    Go to **All Tools** > **AI Agents** > **Brand Voice**, or open [easy-peasy.ai/voices](https://easy-peasy.ai/voices). Click **Add new**, or **Create Your First Brand Voice** if you don't have one yet.

    <Frame>
      <img alt="The Brand Voice page with no voices yet: Train AI to write like you, three steps (Share Your Content, AI Analyzes Style, Generate On-Brand Content), and a Create Your First Brand Voice button" />
    </Frame>
  </Step>

  <Step title="Share your content">
    Choose how to show your style:

    | Option | What to give | What's read |
    | - | - | - |
    | **Import from Website** | Your website's address | Up to 5 pages, such as your home and About pages |
    | **Paste Text Content** | Writing samples: posts, newsletters, product descriptions | The first 3,000 characters |
    | **Upload Files** | Up to 5 TXT, MD, PDF, or DOCX files | Up to 5,000 characters per text file; PDF and DOCX files can be up to 3 MB together |
    | **Describe Your Style** | A description of how you want to write | All of it |

    <Frame>
      <img alt="Create Your Brand Voice with Import from Website selected and the website example-yoga-demo.easy-peasy.site entered, above a How it works box" />
    </Frame>

    Pick samples that sound the way you want the AI to sound. A few good paragraphs work better than a long mix of styles.
  </Step>

  <Step title="Click Continue and name the voice">
    Enter a **Voice Name**, such as "Blog" or your business name. You'll pick the voice by this name later. The **Content Summary** shows what will be analyzed.

    <Frame>
      <img alt="Add details to your brand voice with Voice Name Example Yoga Studio and a Content Summary: Method Website scraping, Website example-yoga-demo.easy-peasy.site" />
    </Frame>
  </Step>

  <Step title="Click Create Brand Voice">
    Your content is analyzed and the writing instructions are written, which takes up to a minute. Then the voice opens.
  </Step>
</Steps>

### Describe your style instead

If you have no samples, choose **Describe Your Style**, pick a starting point, and write a few sentences:

<Frame>
  <img alt="Describe your style: starting points Describe generally, Define style objective, Tailor to an audience, Use specific voice & tone, and Use custom instructions (advanced), with a text box" />
</Frame>

| Starting point | What to write |
| - | - |
| **Describe generally** | Your overall writing style, such as "Clear and direct, short sentences, simple words" |
| **Define style objective** | What your writing should achieve |
| **Tailor to an audience** | Who you write for |
| **Use specific voice & tone** | The voice and tone you want |
| **Use custom instructions** | Finished instructions. They're saved exactly as you write them, without being rewritten |

With the first four, your description is turned into full writing instructions.

## What you get

The voice's page shows its **Instructions for AI**, organized by topic: tone and personality, language style, sentence structure, vocabulary, point of view, how to structure content, and brand personality traits. They often include example phrases to use and to avoid.

<Frame>
  <img alt="Your Brand Voice: Voice Name Example Yoga Studio and Instructions for AI, starting with Writing style instructions and a Tone & Personality section: warm and welcoming, calm and reassuring, inclusive and encouraging" />
</Frame>

Your original samples aren't kept; only the instructions are.

## Edit a brand voice

Open **Brand Voice** and click **Edit** next to a voice.

* To change the instructions, click **Edit** next to **Instructions for AI**, change the text, and click **Save Changes**. Add rules the analysis missed, such as "Never use exclamation marks" or "Always write 'sign up', not 'signup'".
* To rename it, change **Voice Name** and click **Save**.

<Frame>
  <img alt="Editing Instructions for AI in a text box, with Save Changes and Cancel, and Go Back, Delete, and Save at the bottom" />
</Frame>

<Tip>
  Tools read only the first 2,000 characters of the instructions, so put your most important rules at the top. Marky reads all of them.
</Tip>

## Delete a brand voice

Open the voice, click **Delete** at the bottom, and confirm. Deleted voices can't be restored.

## Share with your team

On a Teams plan, the brand voices your team creates appear on everyone's **Brand Voice** page, so the whole team writes the same way.

<Columns>
  <Card title="Use a brand voice" icon="bullhorn" href="/docs/brand-voice/use">
    Write in your voice with Marky and the tools.
  </Card>

  <Card title="Marky Agent" icon="sparkles" href="/docs/marky/overview">
    Everything Marky can write for you.
  </Card>
</Columns>


# Use a brand voice
Source: https://easy-peasy.ai/docs/brand-voice/use

Write in your brand voice with Marky and the writing tools, and switch it off when you don't need it.

Once you've [created a brand voice](/docs/brand-voice/create), Marky and the writing tools can follow its instructions. The voice changes how the text sounds, not what it says: facts, names, and details still come from your request.

## In Marky

<Steps>
  <Step title="Pick the voice">
    In the message box, click **+** > **Brand Voice** and pick your voice.

    <Frame>
      <img alt="Marky's + menu with Brand Voice open, listing None, Example Yoga Studio, and Create a brand voice" />
    </Frame>
  </Step>

  <Step title="Check it's on">
    A megaphone button appears next to **+**. Hover over it to see which voice is on, or click it to switch voices.

    <Frame>
      <img alt="The message box with a request typed, and the megaphone button next to + showing the tooltip Example Yoga Studio" />
    </Frame>
  </Step>

  <Step title="Send your request">
    Marky writes in your voice.

    <Frame>
      <img alt="Marky's welcome email for a new student, with the subject See you on the mat soon, a short list of what to bring, and the lines Every level, every body and Come find us, signed The Example Yoga Studio team" />
    </Frame>
  </Step>
</Steps>

The voice stays on for everything you send, in this chat and in new ones, until you choose **None** or reload the page. Name your business in the request when it matters: the voice sets the style, and Marky may take names from your [memory](/docs/marky/memory) otherwise.

## In the writing tools

More than 130 tools have a **Tone of voice** field, among them Blog Post, LinkedIn Post Generator, and Instagram Post Caption. They can all use your brand voices.

<Frame>
  <img alt="The Instagram Post Caption tool with Tone of voice set to the brand voice Example Yoga Studio, and a Manage voices link" />
</Frame>

* Once you have a brand voice, **Tone of voice** is set to your newest one. Choose another voice from the list.
* To describe a tone for one result instead, choose **Custom** and type it, such as "witty" or "professional".
* **Manage voices** opens your brand voices in a new tab.

<Frame>
  <img alt="An Instagram caption written in the brand voice: Early mornings just got a whole lot softer. Our new Sunrise Flow starts Monday at 6:30 am - we'll save you a spot." />
</Frame>

Tools read the first 2,000 characters of the voice's instructions, so keep the most important rules at the top.

Brand voices aren't available in workflows or on the public tool pages; there, type the tone you want.

## Cost

A brand voice doesn't cost anything extra. Marky and the tools use words as usual for what they write. See [Plans and credits](/docs/account/plans-and-credits#words).

## Tips

* **One voice per audience.** A voice for customers and one for investors work better than one voice that tries to cover both.
* **Fix what you notice.** When results keep doing something you don't want, add a rule to the instructions, such as "No emojis" or "Never promise delivery dates".
* **Give facts in the request.** Prices, dates, and names belong in what you ask for, not in the voice.

<Columns>
  <Card title="Create a brand voice" icon="bullhorn" href="/docs/brand-voice/create">
    Build one from your website or samples.
  </Card>

  <Card title="Prompts and brand voice in Marky" icon="book" href="/docs/marky/prompts-and-brand-voice">
    Reuse good prompts with your voice.
  </Card>
</Columns>


# Chat widget
Source: https://easy-peasy.ai/docs/developers/chat-widget

Embed an agent's chat bubble, configure it with data attributes, and control it from JavaScript with window.easyPeasyChat.

The chat widget puts an [agent](/docs/agents/overview) on any web page. It's one script tag, configured with data attributes, and a small JavaScript API on `window.easyPeasyChat` for opening, hiding, and identifying the visitor. To get the snippet without code, see [Add your agent to your website](/docs/agents/website).

## Embed

```html theme={null}
<script
  src="https://bots.easy-peasy.ai/chat.min.js"
  data-chat-url="https://bots.easy-peasy.ai/bot/YOUR_AGENT_ID"
  data-btn-position="bottom-right"
  defer>
</script>
```

Load it as a classic `<script>` tag, not a module. Your agent ID is under **Channels** > **Website** > **Your Agent ID**.

| Attribute | Values |
| - | - |
| `data-chat-url` | `https://bots.easy-peasy.ai/bot/YOUR_AGENT_ID`. Required |
| `data-btn-position` | `bottom-right` (default), `bottom-left`, `top-right`, `top-left`, `side-right`, `side-left`. A position saved in **Appearance** > **Chat bubble** takes priority |
| `data-widget-btn-color` | The bubble color, such as `#6366f1`. A color saved in **Appearance** > **Chat bubble** takes priority |
| `data-btn-size` | `medium`, `large`, or `xlarge` (corner positions). A saved size takes priority |
| `data-btn-label` | The text on the side tab (side positions). Saved text takes priority |
| `data-widget-icon` | URL of an image, or a built-in icon name: `spark`, `bubble`, `typing`, `chats`, `message`, `smile`, `heart`, `question`, `sparkles`, `bot`, `headset`, `outline`, `classic`. An icon saved in **Appearance** > **Chat bubble** takes priority |
| `data-widget-width` | Width of the open chat on desktop, in pixels, from 320 to 720. Phones always use the full screen. A saved width takes priority |

Apart from `data-chat-url`, you don't need any of these attributes: the bubble's position, size, chat width, icon, color, and text are saved with the agent in [**Appearance** > **Chat bubble**](/docs/agents/appearance#chat-bubble). The script reads them from `https://bots.easy-peasy.ai/bot/YOUR_AGENT_ID/launcher` when the page loads and remembers them in the browser for the next visit. If your site has a Content Security Policy, allow that host in `connect-src`; without it, the bubble uses the attributes above and the default look.

## JavaScript API

The script creates `window.easyPeasyChat`. You can call its methods before the widget has loaded; the calls wait and run once it's ready.

| Method | What it does |
| - | - |
| `open()`, `close()`, `toggle()` | Open or close the chat |
| `hide()`, `show()` | Hide or show the bubble and the chat. The conversation is kept, and `show()` reopens a chat that was open |
| `destroy()`, `init()` | Remove the widget from the page, or create it again. `init()` does nothing while the widget is on the page |
| `isOpen()`, `isVisible()`, `isMounted()` | Current state |
| `setContact(contact)` | Tell the agent who the visitor is. See below |
| `addEventListener(type, fn)`, `removeEventListener(type, fn)` | Listen for `ready`, `open`, `close`, `show`, `hide`, and `destroy` |

When the widget is ready, the window also receives a DOM event, `easy-peasy-chat:ready`.

```html theme={null}
<button onclick="window.easyPeasyChat.open()">Chat with us</button>
```

### Start hidden

To load the widget without showing the bubble, for example on a checkout page, set this before the script tag and call `show()` when you want it:

```html theme={null}
<script>window.easyPeasyChat = { hidden: true };</script>
```

### Single-page apps

Load the script once, in your root layout, and call `hide()` and `show()` on route changes. The widget tells the agent about each client-side navigation, so [Conversations](/docs/agents/conversations#conversation-details) shows the page each chat started on and the last page the visitor saw.

## Identify the visitor

If visitors sign in to your website, pass their details so their conversations show who they are in [Conversations](/docs/agents/conversations):

```js theme={null}
window.easyPeasyChat.setContact({
  email: 'jane@example.com',
  name: 'Jane Doe',
  phone: '+15551234567',
  externalId: 'user_123', // your own ID for the visitor
});
```

Call `setContact(null)` when the visitor signs out. You can also set it before the script loads:

```html theme={null}
<script>
  window.easyPeasyChat = { contact: { email: 'jane@example.com', name: 'Jane Doe' } };
</script>
```

These details label conversations; they aren't verified. Anyone who can run code on your page can set them, so don't use them to decide what the agent may share.

## Iframe

To place the chat inside a page instead of floating over it:

```html theme={null}
<iframe
  src="https://bots.easy-peasy.ai/bot/YOUR_AGENT_ID?mode=embedded"
  width="100%"
  height="600"
  allow="microphone">
</iframe>
```

`allow="microphone"` lets visitors use voice input and [voice calls](/docs/agents/voice).

## Allowed domains and Content Security Policy

If the agent has [**Allowed Domains**](/docs/agents/settings#allowed-domains) set, the widget works only on those domains. If your site sends a Content Security Policy, allow `https://bots.easy-peasy.ai` in `script-src`, `frame-src`, and `connect-src`.

<Columns>
  <Card title="Add your agent to your website" icon="window-maximize" href="/docs/agents/website">
    Build the snippet in the dashboard.
  </Card>

  <Card title="Agents API" icon="robot" href="/docs/api-reference/endpoint/chat-with-bot">
    Talk to your agent from your backend.
  </Card>
</Columns>


# Developer overview
Source: https://easy-peasy.ai/docs/developers/overview

Build on Easy-Peasy.AI with the REST API, the MCP server, and the OpenAI-compatible chat endpoint.

Everything you can create in the Easy-Peasy.AI dashboard is also available to your own code and AI agents. They all use your account's API key and draw from the same plan allowance as the dashboard.

<Columns>
  <Card title="REST API" icon="square-terminal" href="/docs/api-reference/introduction">
    Generate text, images, videos, speech, and transcriptions from your backend.
  </Card>

  <Card title="MCP server" icon="plug" href="/docs/mcp/introduction">
    Connect Claude, Cursor, Windsurf, and other MCP clients without writing API code.
  </Card>

  <Card title="Chat Completions" icon="messages" href="/docs/api-reference/endpoint/chat-completions">
    Point the OpenAI SDK at Easy-Peasy.AI and use the same chat models as Marky.
  </Card>

  <Card title="Agents API" icon="robot" href="/docs/api-reference/endpoint/chat-with-bot">
    Send messages to an AI agent you built in the dashboard.
  </Card>
</Columns>

## Get an API key

Open [**Settings** > **API/MCP**](https://easy-peasy.ai/settings/api) in the dashboard and create a key. Send it in the `x-api-key` header. Keep keys on your server, never in browser code or public repositories.

## What the API covers

| Area | What you can do | Start here |
| - | - | - |
| Text | Run the dashboard's writing tools, from blog posts to SOAP notes | [Generate text](/docs/api-reference/endpoint/generate) |
| Images | Generate and edit images with Nano Banana, GPT Image, FLUX, Seedream, and more | [Generate image](/docs/api-reference/endpoint/generate-image) |
| Video | Text-to-video, image-to-video, and reference-to-video with the dashboard's video models | [Video models](/docs/api-reference/video-models) |
| Talking videos | Animate a face with your script or audio | [Generate talking video](/docs/api-reference/endpoint/generate-talking-video) |
| Audio | Text to speech, sound effects, and transcription | [Text to speech](/docs/api-reference/endpoint/generate-tts) |
| Chat | OpenAI-compatible chat completions with streaming | [Chat Completions](/docs/api-reference/endpoint/chat-completions) |

<Note>
  In the dashboard, the text generators are called **tools**. In the API they're called **presets**, because that's the name of the request field (`preset`).
</Note>

## Quick example

```bash theme={null}
curl -X POST https://easy-peasy.ai/api/generate \
  -H "Content-Type: application/json" \
  -H "x-api-key: YOUR_API_KEY" \
  -d '{
    "preset": "blog-post",
    "keywords": "How to get started with AI",
    "tone": "professional",
    "outputs": 1
  }'
```

API calls use the same allowance as the dashboard. See [Plans and credits](/docs/account/plans-and-credits) for how usage is counted.


# Find your way around
Source: https://easy-peasy.ai/docs/get-started/dashboard

What's on the Home screen, in the sidebar, in History, and in Settings.

The dashboard is where you'll spend most of your time. Here's what each part does.

<Frame>
  <img alt="The dashboard: sidebar on the left, the Home prompt box, Recommended Tools, and Get Inspired templates" />
</Frame>

## Home

The Home screen has three parts:

* **The prompt box.** Anything you type here goes to [Marky Agent](/docs/marky/overview), which opens a chat with the result. The buttons below it (**Create slides**, **Build website**, **Images**, **Music**, **Video**, **More**) start those jobs directly.
* **Recommended Tools.** Popular tools, one click away. Use the arrows to browse, or the magnifier to search every tool.
* **Get Inspired.** Starting points for your own work. **Templates** are ready-made image and video templates you can reuse with your own photos or text, filtered by category. **Image Tools** are one-click image jobs such as **Replace Background**, **Style Transfer**, and **Headshot Generator**. **Examples** shows images other people chose to share publicly.

## The sidebar

<Frame>
  <img alt="The dashboard sidebar with Home, All Tools, favourite tools, History, Settings, Help & Support, and Logout" />
</Frame>

**Home** and **All Tools** are always at the top. Below them are your favourite tools. New accounts start with the main products (Marky Agent, Agents, Workflows, AI Images, AI Videos, and more) already there.

* **To add a tool to the sidebar**, click the star on its card in **All Tools** or on the Home screen.
* **To remove one**, hover over it in the sidebar and click its star.

**History**, **Settings**, **Help & Support**, and **Logout** are always at the bottom.

## The icons under the logo

| Icon | What it does |
| - | - |
| Globe | Changes the dashboard language |
| Sun | Switches between light, dark, and system theme |
| Bell | Opens **What's New**, the list of recent product updates |
| Person | Opens your account menu |

<Frame>
  <img alt="The account menu: My Profile, My Team, Subscription, Usage, History, Favourites, Refer a friend, Help & Support, and Logout" />
</Frame>

## History

**History** keeps everything you create. It's split into tabs: **Text** (tool results and Marky conversations), **Favorites**, **Images**, **Videos**, **AI Text-to-Speech**, **AI Audio Generator**, **Sound Effects**, **Music**, and **Trash**. Use the search box to find something by its text.

Deleted text results go to **Trash** first, so you can restore them. Deleted images and videos can't be restored. See [History and trash](/docs/get-started/history).

## Settings

| Tab | What's there |
| - | - |
| **Profile** | Your name, email address, and password. You can also delete your account here. See [Profile and preferences](/docs/account/profile). |
| **Team** | Seats, invitations, and each member's usage. See [Teams](/docs/account/team). |
| **Interface** | Dashboard language and theme. |
| **Subscription** | Your plan, payment details, billing history, and extra credits. See [Manage your subscription](/docs/account/subscription). |
| **Usage** | How much of this period's words and credits you've used. See [Check your usage](/docs/account/usage). |
| **Data Controls** | Import your ChatGPT conversations, and manage the websites, apps, and domains Marky built for you. See [Data controls](/docs/account/data-controls). |
| **API/MCP** | API keys and the MCP server connection. See [Developers](/docs/developers/overview). |
| **Refer a friend** | Your referral link and bonuses. See [Refer a friend](/docs/account/referrals). |

## Getting help

**Help & Support** in the sidebar has answers to common questions, a live chat, and our email address, [support@easy-peasy.ai](mailto:support@easy-peasy.ai).


# History and trash
Source: https://easy-peasy.ai/docs/get-started/history

Find everything you've created, keep favourites, and restore or permanently delete what you've thrown away.

**History** in the sidebar keeps what you create, sorted by type and newest first.

<Frame>
  <img alt="The History page with the tabs Text, Favorites, Images, Videos, AI Text-to-Speech, AI Audio Generator, Sound Effects, Music, and Trash, and a search box" />
</Frame>

| Tab | What's there |
| - | - |
| **Text** | Results from [tools](/docs/tools/overview) and Marky's replies, 50 per page. **Open Chat** on a reply goes back to its conversation. Search finds items by their text |
| **Favorites** | Results you marked with **Favorite** |
| **Images**, **Videos** | Everything from [AI Images](/docs/ai-images/overview) and [AI Videos](/docs/videos/overview) |
| **AI Text-to-Speech**, **AI Audio Generator**, **Sound Effects**, **Music** | Your [audio](/docs/audio/text-to-speech) |
| **Trash** | Text items you deleted |

Each item has the same buttons as where you made it: copy, save, share, and so on. On a Teams plan, the account owner can filter **Images** and **Videos** by team member.

## Restore or empty the trash

Deleted text results go to **Trash**. Click the restore icon on one to move it back.

**Empty Trash** permanently deletes the content of everything in the trash. It can't be undone. The words those results used still count in your usage for the period.

<Note>
  Only text results go to the trash. Deleted images, videos, songs, sound effects, and speech are deleted straight away and can't be restored, and deleting doesn't give back the credits or characters they used.
</Note>

**Favorites** and **Trash** show your 200 most recent items.

<Columns>
  <Card title="Find your way around" icon="compass" href="/docs/get-started/dashboard">
    The Home screen, sidebar, and Settings.
  </Card>

  <Card title="Check your usage" icon="chart-simple" href="/docs/account/usage">
    See what you've used this period.
  </Card>
</Columns>


# Quickstart
Source: https://easy-peasy.ai/docs/get-started/quickstart

Create an account, give Marky a task, run a tool, and find your results.

This walkthrough takes about five minutes. You'll ask Marky Agent for something, run one of the tools, and then find both results in your history.

<Steps>
  <Step title="Create your account">
    Go to [easy-peasy.ai](https://easy-peasy.ai/login) and sign up with Google, Facebook, Apple, or your email address. If you use email, open the confirmation link we send you.

    New accounts start on the Free plan, so you can try everything on this page before choosing a plan. See [Plans and credits](/docs/account/plans-and-credits) for what's included.
  </Step>

  <Step title="Give Marky a task">
    The Home screen opens with one question: **How can I help you today?** Type what you need in the box and press **Enter**.

    <Frame>
      <img alt="The Home prompt box with a request typed in: write a welcome email for a yoga studio" />
    </Frame>

    Marky opens a chat and writes the answer. Keep the conversation going to change the result: "make it shorter", "add a line about parking", "translate it to Spanish".

    <Frame>
      <img alt="Marky's reply in the chat: a welcome email for Sunrise Yoga with a subject line, greeting, and sign-off" />
    </Frame>

    <Tip>
      The buttons under the prompt box start common jobs in one click: **Create slides**, **Build website**, **Images**, **Music**, **Video**, and **More**.
    </Tip>
  </Step>

  <Step title="Run a tool">
    Tools are single-purpose generators with a short form to fill in. Click **All Tools** in the sidebar, then search or pick a category.

    <Frame>
      <img alt="The Tools page with a search box, category filters such as Writing, Social Media, and Marketing, and tool cards" />
    </Frame>

    Open a tool, fill in the fields, and click **Generate**. This example uses **Instagram Post Caption**. Each result has buttons to **Edit**, **Favorite**, **Share**, **Save**, and **Copy** it. See [Run a tool](/docs/tools/run).

    <Frame>
      <img alt="A generated Instagram caption under the Generate button, with Edit, Favorite, Share, Save, and Copy buttons" />
    </Frame>
  </Step>

  <Step title="Find your results">
    Everything you create is saved. Click **History** in the sidebar to see it, sorted into tabs such as **Text**, **Images**, **Videos**, and **Music**. Marky conversations appear here too; click **Open Chat** to continue one.

    <Frame>
      <img alt="The History page listing the Instagram caption and the Marky Agent email" />
    </Frame>
  </Step>
</Steps>

## Next steps

<Columns>
  <Card title="Find your way around" icon="compass" href="/docs/get-started/dashboard">
    Tour the sidebar, the Home screen, and Settings.
  </Card>

  <Card title="Plans and credits" icon="coins" href="/docs/account/plans-and-credits">
    See how words and credits are counted.
  </Card>

  <Card title="Marky Agent" icon="sparkles" href="/docs/marky/overview">
    Everything Marky can do, from research to slides and websites.
  </Card>

  <Card title="Tools" icon="toolbox" href="/docs/tools/overview">
    Hundreds of single-purpose writing tools, and your own.
  </Card>

  <Card title="AI Images" icon="image" href="/docs/ai-images/overview">
    Create and edit images with dozens of models.
  </Card>

  <Card title="AI Videos" icon="video" href="/docs/videos/overview">
    Generate, edit, and extend videos, or make one from a sentence.
  </Card>

  <Card title="Agents" icon="robot" href="/docs/agents/overview">
    Train an agent on your content and put it on your website and channels.
  </Card>

  <Card title="Audio" icon="waveform-lines" href="/docs/audio/text-to-speech">
    Speech, transcripts, dubbing, music, and sound effects.
  </Card>
</Columns>


# Welcome to Easy-Peasy.AI
Source: https://easy-peasy.ai/docs/index

One workspace for an AI agent, image and video generation, voice and audio tools, and hundreds of focused tools.

<Frame>
  <img alt="The Easy-Peasy.AI dashboard: a prompt box that asks 'How can I help you today?', recommended tools, and templates to get inspired" />
</Frame>

Easy-Peasy.AI brings the AI you use for work into one place. Ask Marky Agent to handle a task, generate images and videos with current models, turn scripts into speech, transcribe recordings, or build an AI agent that answers your customers on your website and messaging apps. Everything runs on one account and one plan.

<Card title="Quickstart" icon="rocket" href="/docs/get-started/quickstart">
  Give Marky a task, run a tool, and find your results in about five minutes.
</Card>

## What you can do

<Columns>
  <Card title="Marky Agent" icon="sparkles" href="/docs/marky/overview">
    An AI agent that writes, researches, builds slides and websites, and works with your connected apps.
  </Card>

  <Card title="AI Images" icon="image" href="/docs/ai-images/overview">
    Generate and edit images with models like GPT Image, Nano Banana, FLUX, and Seedream.
  </Card>

  <Card title="AI Videos" icon="video" href="/docs/videos/overview">
    Turn text, images, or reference clips into video, then extend, edit, or upscale it.
  </Card>

  <Card title="Voice and audio" icon="waveform-lines" href="/docs/audio/text-to-speech">
    Text to speech, voice cloning, transcription, music, and sound effects.
  </Card>

  <Card title="Tools" icon="wand-magic-sparkles" href="/docs/tools/overview">
    Focused tools for emails, social posts, lesson plans, clinical notes, and hundreds of other jobs.
  </Card>

  <Card title="Agents" icon="robot" href="/docs/agents/overview">
    Train an AI agent on your content and put it on your website, WhatsApp, Slack, and more.
  </Card>
</Columns>

## Start here

<Columns>
  <Card title="Find your way around" icon="compass" href="/docs/get-started/dashboard">
    What's in the sidebar, on the Home screen, in History, and in Settings.
  </Card>

  <Card title="Plans and credits" icon="coins" href="/docs/account/plans-and-credits">
    How words and credits work, and what each plan includes.
  </Card>

  <Card title="Build with the API" icon="code" href="/docs/developers/overview">
    Use the REST API, the MCP server, and OpenAI-compatible chat from your own code.
  </Card>

  <Card title="Get help" icon="life-ring" href="https://easy-peasy.ai/support">
    Browse the FAQ, chat with support, or email [support@easy-peasy.ai](mailto:support@easy-peasy.ai).
  </Card>
</Columns>


# Browse the web
Source: https://easy-peasy.ai/docs/marky/browser

Let Marky use websites for you: in a browser in the cloud, or in your own Chrome with your logins.

Some jobs need a real browser: reading a page that search can't open, filling in a form, collecting data from a site you're logged in to, or checking a website Marky just built. Marky can use two browsers:

| Browser | How it works | Best for |
| - | - | - |
| **Cloud browser** | A browser Marky runs in the cloud. You watch it in **Marky's Computer**. | Public websites, research, testing sites. Works out of the box. |
| **My Browser** | Your own Chrome, through the Easy-Peasy.AI Browser Agent extension. Marky works in a tab you can see. | Sites where you're already logged in, such as LinkedIn, your CRM, or admin panels. |

Just describe the job: "Go to example.com and list every product under \$50", "Check the contact form on my new site works."

## The cloud browser

Marky opens pages, clicks, types, scrolls, and reads what's on screen. After each action, a screenshot appears in **Marky's Computer**, so you can follow along.

Logins you make in the cloud browser are saved for your account, so Marky stays logged in across chats. You can turn this off when you finish a takeover.

### Take over the browser

When Marky reaches a login page, a verification code, or a payment, it doesn't guess. It asks you to take over:

1. In the chat, click **Take control** on the **Marky needs you in the browser** card. Click **I can't right now** to skip.
2. The browser opens full screen. Log in, or do what Marky asked. What you type stays in the browser and isn't shared with the AI.
3. Click **Exit takeover**, or press **Esc**.
4. Tell Marky what you did, if anything needs explaining, and click **Send and continue**. Turn off **Persist login state across tasks** if you don't want the login kept for later chats.

## Use your own browser

With the Easy-Peasy.AI Browser Agent extension, Marky works in your Chrome, with your logins and your internet connection. It works in Chrome, Edge, Brave, and Arc.

<Steps>
  <Step title="Install the extension">
    Open **Connectors** in the Marky sidebar and click **Install** next to **My Browser**, or install [Easy-Peasy.AI Browser Agent](https://chromewebstore.google.com/detail/easy-peasyai-browser-agen/mkigjcomdaidhcnpggmajpbbbcdppbcf) from the Chrome Web Store.
  </Step>

  <Step title="Connect this browser">
    Open [easy-peasy.ai/browser-extension](https://easy-peasy.ai/browser-extension) in the same browser, signed in to your account. Read the disclosure, check the box, and click **Connect this browser**. When it's connected, **My Browser** shows **Online** in **Connectors**.

    <Frame>
      <img alt="The Set up in two steps section: Install the extension, with Open the Web Store and Detected in this browser, and Connect this browser, with a disclosure checkbox about what is sent to Easy-Peasy.AI and a Connect this browser button" />
    </Frame>
  </Step>

  <Step title="Ask Marky to use it">
    In any chat, give Marky a job on a website. While your browser is online, Marky uses it by default. It opens a new tab in your window, in a tab group named **Marky**, and works there while you watch. You can click into the tab at any time; Marky waits and picks up where you left off.
  </Step>
</Steps>

### Choose which sites Marky can use

The first time Marky wants to work on a website, the extension's window opens and asks: **Allow Marky to work on** the site? If it doesn't open, click the extension icon in your toolbar; it shows an orange **?** while a question is waiting.

* **Allow on** the site: Marky can work there from now on.
* **Just this once**: allows this action only.
* **Deny**: Marky stops and tells you.

If you don't answer within about 90 seconds, the action is cancelled and Marky asks in the chat before trying again.

To review or remove sites, click the extension icon in your browser's toolbar and look under **Sites Marky can work on**. You can also switch on **Allow every site without asking**, or click **Stop and disconnect** to end the session.

<Note>
  Marky asks in the chat before it does anything you can't take back, such as sending a message, posting, or buying something. Keep the browser open while Marky works: the first action can take up to 30 seconds to start.
</Note>

### Turn it off or disconnect

* In **Connectors**, switch **My Browser** off. Your browser stays connected, but Marky uses the cloud browser instead.
* To disconnect a browser, open **My Browser** in **Connectors** and click **Manage browsers**, or open the [browser extension page](https://easy-peasy.ai/browser-extension) and click **Disconnect** next to it.

## Limits

* Marky can't use your browser in [scheduled tasks](/docs/marky/scheduled-tasks) unless your computer is on and Chrome is open.
* Some websites block automated browsers. If a site shows a security check, Marky asks you to take over, or you can use **My Browser** instead.

## Credits

Browsing doesn't cost extra: Marky's work uses words at the rate of the chat's model, like any other task.

<Columns>
  <Card title="Connectors" icon="plug" href="/docs/marky/connectors">
    Give Marky access to your apps through their official connections.
  </Card>

  <Card title="Research" icon="magnifying-glass" href="/docs/marky/research">
    Web search with sources and in-depth reports.
  </Card>
</Columns>


# Chat with Marky
Source: https://easy-peasy.ai/docs/marky/chat

Send Marky a task, attach files, follow its work step by step, and answer its questions.

This page covers the basics of working with Marky: the message box, attaching files, following Marky's work, and what to do when you want to stop or redo something.

## Send a task

<Steps>
  <Step title="Start a chat">
    Click **New Chat** in the Marky sidebar, or type into the prompt box on the dashboard home page.
  </Step>

  <Step title="Describe what you want">
    Type in the box that says **Assign a task or ask anything**. The more context you give, the better the result: who it's for, the tone, the length, and what "done" looks like. For example: "Find three beginner-friendly yoga class formats studios are promoting this year, with a source for each."
  </Step>

  <Step title="Send it">
    Click the send button. By default, **Enter** starts a new line and **Cmd+Enter** (or **Ctrl+Enter** on Windows) sends. You can switch this in [Chat settings](#chat-settings).
  </Step>
</Steps>

The first message in a new chat can take a few seconds longer, while Marky starts its computer.

<Frame>
  <img alt="Marky's answer listing three beginner yoga class formats, each followed by the website it came from, with the search steps and Thought for 4s above it, and the Regenerate, Favorite, Copy, More, and Was this helpful? buttons below it" />
</Frame>

## Attach files

Click **+** > **Add files or photos**, drag files onto the message box, or paste them. Marky can read documents (PDF, Word, PowerPoint, Excel, CSV, text), images, archives, code, subtitles, audio, and video.

* Up to 100 files per chat.
* Up to 50 MB per file on the Free plan and 500 MB on paid plans. Videos can be up to 500 MB on every plan.
* Wait for the upload to finish before you send.

Files are copied to Marky's computer for that chat, where it can open, analyze, and convert them. Ask for what you need: "summarize this contract", "make a chart from this spreadsheet", "turn this PDF into slides".

<Note>
  Only models with **Code Interpreter** can read files. If the current model can't, you'll see a message asking you to switch. See [Models and words](/docs/marky/models).
</Note>

## Speak instead of typing

Click the microphone, speak for up to 60 seconds, and stop. Marky adds the transcribed text to the message box so you can check it before sending.

## The + menu

<Frame>
  <img alt="The + menu open above the message box, with Add files or photos, Brand Voice, Skills, Prompts, and Connectors" />
</Frame>

| Item | What it does |
| - | - |
| **Add files or photos** | Attaches files to the chat. |
| **Brand Voice** | Writes in the tone of one of your brand voices. See [Prompts and brand voice](/docs/marky/prompts-and-brand-voice). |
| **Skills** | Puts one of your skills to work on this message. See [Skills](/docs/marky/skills). |
| **Prompts** | Inserts one of your saved prompts. |
| **Connectors** | Shows which apps are connected and on. See [Connectors](/docs/marky/connectors). |

## Follow Marky's work

As Marky works, each step appears in the chat: searches, commands, files it reads and writes. Click a step to see it in detail.

* **Thinking**: some models think before they answer. Click **Thought for Xs** to read their reasoning.
* **Task progress**: for longer jobs, Marky writes a plan. The bar above the message box shows the current phase and how many are done. Click it to see the whole checklist.
* **Marky's Computer**: a panel on the right that shows what Marky is doing, such as search results, commands and their output, websites it opens, and files it writes. Use the arrows or the slider at the bottom to step through earlier actions, and **Jump to latest** to follow along again. The monitor icon at the top of the chat opens and closes the panel.

<Frame>
  <img alt="The Marky's Computer panel on the right, showing Marky is using Search with 15 web results for a query about beginner yoga classes, next to Marky's answer in the chat" />
</Frame>

## When Marky asks you something

Sometimes Marky needs a decision before it continues, such as approving a style or choosing between options. It shows a question card with choices. Pick one, or choose **Other** and type your own answer, then click **Submit answers**.

If Marky needs you to log in to a website or complete a payment in its browser, it asks you to take over. See [Browse the web](/docs/marky/browser#take-over-the-browser).

## Stop, redo, or change a message

* **Stop**: while Marky works, the send button becomes a stop button. Marky stops after the current step and keeps what it has done so far.
* **Regenerate**: under Marky's latest reply, click **Regenerate** to get a new answer to your last message.
* **Edit**: hover over one of your messages and click **Edit** to change it. Marky's replies after it are removed, and it answers the new version.
* **Continue**: if you see **Marky stopped before finishing**, click **Continue** to pick up where it left off.

Marky keeps working when you close the tab or switch chats. When you come back, the reply is there.

## Reply actions

Under each reply you'll find **Regenerate**, **Favorite**, and **Copy**, and in the **⋯** menu, **Read Aloud** and **Fork chat** (a copy of the conversation up to that reply, in a new chat). The thumbs up and down under **Was this helpful?** send us feedback about the answer.

## Chat settings

Click the gear icon at the top of the chat.

<Frame>
  <img alt="The Settings window with Memory, Model, Chat Persona, Image Model, Press Enter to, Open Marky's Computer automatically, and Should AI speak back? settings" />
</Frame>

| Setting | What it does |
| - | - |
| **Memory** | Opens your saved memories. See [Memory](/docs/marky/memory). |
| **Model** | The AI model for this chat. See [Models and words](/docs/marky/models). |
| **Chat Persona** | Gives Marky a role, such as **Coding Assistant**, **Travel Guide**, or **Legal Advisor**. The default is **General AI**. |
| **Image Model** | The model Marky uses when it makes images in chat. See [Images, music, and video](/docs/marky/media). |
| **Press Enter to** | **Insert new line** (the default; send with Cmd/Ctrl+Enter) or **Send message** (Shift+Enter for a new line). |
| **Open Marky's Computer automatically** | Opens the panel whenever Marky uses a tool. Turn it off to open the panel only when you want. |
| **Should AI speak back?** | Reads Marky's replies aloud. This is in beta. |

<Columns>
  <Card title="Manage your chats" icon="folder-open" href="/docs/marky/manage-chats">
    Search, organize in folders, share, and delete chats.
  </Card>

  <Card title="Models and words" icon="microchip" href="/docs/marky/models">
    Choose the model for a chat and see what it costs.
  </Card>
</Columns>


# Connectors
Source: https://easy-peasy.ai/docs/marky/connectors

Connect Gmail, Google Calendar, Notion, GitHub, Stripe, and dozens of other apps so Marky can read and act in them.

Connectors give Marky access to the apps you work in. With Gmail connected, you can ask "summarize the unread emails from clients and draft replies". With Google Calendar: "find a free hour with Anna next week and book it". With Stripe: "how many new subscriptions did we get this month?"

Marky reads your data only when a task needs it. In email, calendar, file, and note apps, it checks with you before it sends, creates, changes, or deletes anything.

## Connect an app

<Steps>
  <Step title="Open Connectors">
    Click **Connectors** in the Marky sidebar. You can also click **+** in the message box, then **Connectors** > **Manage Connectors**.

    <Frame>
      <img alt="The Connectors window with the Apps, Custom API, and Custom MCP tabs, a search box, and rows for My Browser, Gmail, Google Calendar, Google Drive, GitHub, Notion, and Outlook Mail, each with a Connect or Install button" />
    </Frame>
  </Step>

  <Step title="Click Connect">
    Find the app in the list or search for it, and click **Connect**. Sign in to the app in the window that opens and approve access.
  </Step>

  <Step title="Use it in a chat">
    Back in Marky, the app shows **Connected** with the account you used. Ask Marky to do something with it. Connected apps also appear as small logos next to **+** in the message box.
  </Step>
</Steps>

## Available apps

| Category | Apps |
| - | - |
| Email and calendar | Gmail, Google Calendar, Outlook Mail, Outlook Calendar, Cal.com |
| Files and notes | Google Drive, OneDrive, OneNote, Notion, Airtable |
| Work management | Asana, ClickUp, Linear, monday.com, Todoist, Trello |
| Sales and support | Attio, Close, Crisp |
| Payments | PayPal, Square, Stripe |
| Marketing | Customer.io, Klaviyo, MailerLite, Meta Ads, Tally |
| Analytics and SEO | Ahrefs, DataForSEO, Mixpanel, Mixpanel (EU), PostHog, Semrush |
| Websites and development | Cloudinary, GitHub, Neon, Sentry, Supabase (Read-only), Supabase (Read & Write), Vercel, Webflow, Wix |
| Automation | Zapier |
| Your browser | My Browser. See [Browse the web](/docs/marky/browser#use-your-own-browser). |

Some apps come in two versions so you can choose how much Marky may do. For example, **Supabase (Read-only)** lets Marky look at your database but never change it.

For **Crisp**, you paste a token instead of signing in: in Crisp, go to **Workspace Settings** > **Advanced** and copy the **MCP Server Token**.

Don't see your app? Click **Let us know!** at the bottom of the window, or add it yourself as a [custom connector](#add-a-custom-connector).

## Turn connectors on and off

Each connected app has a switch. Turn it off and Marky won't use that app in any chat, but it stays connected, so you can turn it back on in one click. This is useful when you want Marky to stay out of your email for a while.

To see what Marky can do in an app, click the app's name. The details list its **Tools**, such as searching, reading, or creating. **Disconnect** removes the connection.

## Add a custom connector

If an app isn't in the list, connect it yourself with **Create** at the top of the window:

| Option | Use it for |
| - | - |
| **Custom API** | Any service with an HTTP API. Enter a **Name** and the **Base URL**, then add your API key under **Secrets**. Marky only sends requests to that address. |
| **Custom MCP** | A service that offers an MCP server. Enter the **Server URL** and any **Custom headers** (for example, an API key), or a **Command** to run the server in Marky's computer. |
| **Import MCP by JSON** | Paste an `mcpServers` configuration from Claude Desktop, Cursor, or another app. |

Custom MCP servers connect with URLs and headers; signing in with OAuth works only for the apps in the list.

## Privacy and Google data

* Marky reads your data only while it works on a task that needs it.
* In Gmail, Outlook, Google Calendar, Outlook Calendar, Google Drive, OneDrive, OneNote, Notion, and GitHub, Marky asks in the chat before it sends an email, changes your calendar, or creates or changes files, pages, and issues. For other apps, say in your request if you want Marky to check with you first, for example "show me the refund before you issue it".
* Grok (xAI) and DeepSeek models can't use Gmail, Google Calendar, or Google Drive, because of Google's rules for how its data may be used. While those connectors are on, these models are unavailable in the model picker. Turn the Google connectors off to use them.

<AccordionGroup>
  <Accordion title="Marky says an app isn't connected, but it is">
    The app's sign-in may have expired, for example after you changed your password or haven't used it for a long time. Open **Connectors** and connect the app again. If it still shows **Connected**, click its name, then **Disconnect**, and connect it again.
  </Accordion>

  <Accordion title="Can my scheduled tasks use connectors?">
    Yes. [Scheduled tasks](/docs/marky/scheduled-tasks) can use every connector that's connected and on.
  </Accordion>

  <Accordion title="Which OneNote notebooks can Marky see?">
    Notebooks stored in your own OneDrive or OneDrive for Business. Notebooks on SharePoint team sites or in Microsoft 365 groups aren't available.
  </Accordion>
</AccordionGroup>

<Columns>
  <Card title="Browse the web" icon="window-maximize" href="/docs/marky/browser">
    For websites without a connector, let Marky use a browser.
  </Card>

  <Card title="Scheduled tasks" icon="calendar-clock" href="/docs/marky/scheduled-tasks">
    Run a connector job every day or week.
  </Card>
</Columns>


# Manage your chats
Source: https://easy-peasy.ai/docs/marky/manage-chats

Find old chats, organize them in folders, share a chat with a public link, and delete chats you don't need.

All your Marky conversations are in the sidebar under **Your Chats**, with favorites first and then the newest. A small spinner next to a chat means Marky is still working on it.

To collapse the sidebar, click the panel icon at the top. **Back to Home** returns to the dashboard.

## Find a chat

Type in **Search...** in the sidebar. It searches chat titles and the text of your messages.

Your Marky chats also appear in **History** > **Text** on the dashboard, where **Open Chat** takes you back to one.

## Organize chats

Hover over a chat and click **⋯** for these options:

| Option | What it does |
| - | - |
| **Favorite** | Pins the chat to the top of the list. |
| **Share** | Creates a public link. See [Share a chat](#share-a-chat). |
| **Move to** | Moves the chat into a folder, or back to **Uncategorized**. |
| **Rename** | Changes the chat's title. |
| **Delete** | Deletes the chat for good. See [Delete a chat](#delete-a-chat). |

To make a folder, click the folder icon next to **Your Chats** and type a name. You can also drag chats onto a folder. Deleting a folder moves its chats to **Uncategorized**; the chats aren't deleted.

## Share a chat

1. Click the share icon at the top of the chat, or **Share** in the chat's **⋯** menu.
2. Turn on **Share Publicly**.
3. Copy the link, or share it straight to LinkedIn, Facebook, X, or Reddit.

Anyone with the link can read the conversation, but can't continue it. Turn **Share Publicly** off to take the link down.

On Teams plans, **Share with Team Members** makes the chat visible to your team instead of the public.

## Copy a conversation

To try a different direction without losing the original, open the **⋯** menu under one of Marky's replies and click **Fork chat**. You get a new chat with everything up to that reply, named after the original with "(forked)".

## Delete a chat

Deleting a chat is permanent and can't be undone. It also removes everything Marky made in that chat:

* The chat's messages and the files on Marky's computer.
* Any website or app built in the chat, including the published site, its database, and its settings.
* Scheduled tasks that post to the chat.

Slides and files you downloaded stay on your computer. Images and videos Marky made stay in **History**.

<Warning>
  If a chat has a published website, deleting the chat takes the site offline. Download the code first with **⋯** > **Download ZIP** in the workspace if you want to keep it. See [Build websites and apps](/docs/marky/websites).
</Warning>

<Columns>
  <Card title="Chat with Marky" icon="message" href="/docs/marky/chat">
    Send tasks, attach files, and follow Marky's work.
  </Card>

  <Card title="Memory" icon="brain" href="/docs/marky/memory">
    Let Marky remember you across chats.
  </Card>
</Columns>


# Images, music, and video in chat
Source: https://easy-peasy.ai/docs/marky/media

Ask Marky to make images, songs, videos, and voiceovers, and use them in the same conversation.

Marky can make media as part of a task: a hero image for the blog post it just wrote, a jingle for an ad, a short product video, or a voiceover for a script. Ask for it in plain language, or start a new chat with the **Images**, **Music**, or **Video** button to see example prompts.

For more control over a single image or video, such as picking from dozens of models, use [AI Images](/docs/ai-images/overview) and [AI Videos](/docs/videos/overview) instead. Everything Marky makes also appears in your **History**.

## Images

Ask for an image and Marky describes it to an image model and shows the result in the chat. Each image has two buttons in its corner: **edit** opens it in the image editor, and **download** saves it.

<Frame>
  <img alt="An image Marky made in the chat, a sunlit yoga studio with rolled mats and plants, with edit and download buttons in the top right corner" />
</Frame>

* Marky uses the **Image Model** set in [Chat settings](/docs/marky/chat#chat-settings). The default is **GPT Image 2.5 Flare Medium**, at 3 image & video credits per image. With **Images** mode on, the model is also shown next to the message box.
* To change an image, ask: "Make the background white", "Same image, but in a vertical 9:16 format".
* To use your own photo, attach it and say what to do with it: "Put this product on a marble kitchen counter."

| Image model in chat | Credits per image |
| - | - |
| Z-Image Turbo | 0.1 |
| GPT Image 2.5 Flare Low, GPT Image 2 Low | 0.5 |
| Nano Banana 2 Lite, Seedream 4.5 | 2 |
| GPT Image 2.5 Flare Medium, GPT Image 2 Medium | 3 |
| Nano Banana 2 | 4 |
| Nano Banana Pro | 7 |
| GPT Image 2.5 Flare High, GPT Image 2 High | 11 |

## Music

Ask for instrumental music or a song: "an upbeat 30-second jingle for a yoga studio", or give Marky lyrics to sing. Marky writes the lyrics if you don't. Songs take about 3 to 4 minutes, and Marky shows a player in the chat when they're ready.

Each song uses 600 text-to-speech characters from your plan. Two versions are made, 300 characters each; Marky plays the first, and both are in [AI Music Generator](/docs/audio/music).

## Video

Ask for a short video, with or without a starting image: "a 10-second clip of a sunrise over a yoga studio, slow camera push-in". Marky can also animate an image from the chat, or keep a character consistent from reference images you attach.

* Videos are 5 to 15 seconds long, with sound.
* They cost 4 image & video credits per second, plus a little extra for reference images. Marky tells you the price when it starts.
* Videos in chat need a paid plan. If a video fails, no credits are charged.

## Speech

Ask Marky to read a text aloud, record a voiceover for a script, or make a short narration, up to 5,000 characters at a time. Speech in chat counts half the length of the text in text-to-speech characters: a 1,000-character script uses 500. To pick from hundreds of voices and fine-tune the delivery, use AI Text to Speech on the dashboard.

## When you're out of credits

Marky tells you when an image, video, or song can't be made because your plan's allowance is used up. Buy more with [bonus credits](/docs/account/buy-credits) or wait for your plan to renew. Your words aren't affected: you can keep chatting.


# Memory
Source: https://easy-peasy.ai/docs/marky/memory

Let Marky remember facts about you and your business across chats, add memories yourself, or import your ChatGPT history.

With memory on, you don't have to repeat yourself. Marky remembers useful facts from your conversations, such as your business, your audience, and how you like things written, and uses them in every new chat.

## Turn memory on

Click the book icon at the top of a chat, or open [Chat settings](/docs/marky/chat#chat-settings) and click **Manage Memories**. Then switch on **Enable Memory**.

<Frame>
  <img alt="The Memory window with Enable Memory switched on, Add new memory and Import buttons, and two saved memories: prefers short, friendly copy without emojis, and runs Sunrise Yoga, a small yoga studio in Austin, Texas" />
</Frame>

From then on, Marky saves a memory when you tell it something worth keeping. You can also ask directly: "Remember that our brand color is sage green."

## Add, remove, and review memories

* **Add new memory**: type a fact, up to 500 characters, and press **Enter**.
* **Your Memories** lists everything Marky remembers, newest first. Delete a memory you don't want with its trash icon, or remove them all with **Delete All**.
* To change a memory, delete it and add the new version.

Marky reads your newest memories first. If you have a very long list, the oldest ones may not fit, so keep memories short and remove ones that are out of date.

## Bring your memory from another AI

Click **Import** in the Memory window. Copy the prompt shown there into the other assistant, such as ChatGPT or Claude, paste its answer back, and click **Add to memory**. Memory turns on automatically.

## Import your ChatGPT conversations

You can move your whole ChatGPT history into Marky: your conversations become Marky chats you can continue, and Marky can learn about you from them.

<Steps>
  <Step title="Export your data from ChatGPT">
    In ChatGPT, open **Settings** > **Data controls** > **Export data**. OpenAI emails you a download link. Download the ZIP file.
  </Step>

  <Step title="Upload it to Easy-Peasy.AI">
    On the dashboard, open **Settings** > **Data Controls** and find **Import from ChatGPT**. Upload the ZIP file (up to 2 GB), or just the `conversations.json` file from it. It's read in your browser, and you see how many conversations, messages, and images it found.
  </Step>

  <Step title="Choose what to import">
    * **Also teach Marky about me** (on by default): Marky reads your custom instructions and the topics of your conversations, and saves up to 30 memories about you. Memory turns on.
    * **Replace conversations I already imported**: turn this on if you're importing again and want the earlier copies updated.
  </Step>

  <Step title="Start the import">
    Your conversations appear in a **ChatGPT** folder in the Marky sidebar, with their original dates.
  </Step>
</Steps>

### What's imported

* The final version of each conversation, as you saw it in ChatGPT. Earlier edits and regenerated answers are skipped.
* Images you uploaded to ChatGPT, up to 25 MB each.
* Not imported: images ChatGPT generated (OpenAI doesn't include them in the export), audio, and files.

Importing is free and doesn't use words. Imported chats are private to you, even on a Teams plan.

<AccordionGroup>
  <Accordion title="Is memory shared with my team?">
    No. Memories belong to your account, and each person on a team has their own.
  </Accordion>

  <Accordion title="Does memory work in every chat?">
    Yes, once it's on, in every Marky chat. Turn **Enable Memory** off to stop Marky from saving and using memories; your saved memories stay until you delete them.
  </Accordion>

  <Accordion title="Can I import the same ChatGPT export twice?">
    Yes. Conversations that are already imported are skipped, so nothing is duplicated. Turn on **Replace conversations I already imported** to update them instead.
  </Accordion>
</AccordionGroup>


# Models and words
Source: https://easy-peasy.ai/docs/marky/models

Choose the AI model for a chat, and see how many words each model uses from your plan.

Marky can run on more than a hundred AI models from OpenAI, Anthropic, Google, xAI, Meta, DeepSeek, and others. Faster, smaller models are cheaper and good for everyday tasks. The biggest models are better at hard reasoning, long documents, and complex builds, and use more of your words.

The default is **GLM 5.3 Flash**: fast, capable, and one of the cheapest to run.

## Choose a model

Click the model name at the top left of the chat.

<Frame>
  <img alt="The model picker with a search for GLM 5.3, four Z.ai models with their word rates, and the details of GLM 5.3 Flash: description, Intelligence Index 57 of 100, a 1,000,000-token context length, and 0.05x word consumption" />
</Frame>

* Search by name, or browse by maker under **Providers**.
* **Leaderboard** ranks the models by their **Intelligence Index**, a combined score for reasoning, knowledge, math, and coding from Artificial Analysis.
* Click a model to see its description, specifications, word rate, and features, then click **Select**.

The model you pick is saved with the chat, and new chats start with the model you used last. Models marked **Pro** need a paid plan.

<Tip>
  Start with the default. Switch to a bigger model when a task needs deeper reasoning, such as a complex website, a long legal document, or tricky analysis. For building websites and apps, Marky suggests **Claude Opus 5**.
</Tip>

## How words are counted

Marky counts the words it writes, including its reasoning and the text of every step of a task. Each model has a **word rate** that scales the count:

* At **1x**, 1,000 words from Marky use 1,000 words from your plan.
* At **0.05x**, the same 1,000 words use 50.
* At **2x**, they use 2,000.

The rate is shown next to each model in the picker and in the table below.

Some large models also count part of what they read, such as long conversations and attached documents. This applies to Claude Opus, Sonnet, and Fable, GPT-6 Astra, GLM 5.3, and Muse Spark 1.3. Their cost grows with the length of the chat, so start a new chat when you switch topics.

### Plan rules

* **Free plan:** every model counts at 0.2x, whatever the picker says.
* **Unlimited** in the picker means these models don't use words on Unlimited and Teams plans. On other paid plans, they count at 0.1x.
* When you run out of words, Marky stops and offers to upgrade. On Unlimited and Teams plans, you can keep chatting with the models that don't count words.

Check your balance in [Usage](/docs/account/usage), and see [Plans and credits](/docs/account/plans-and-credits) for how words work across Easy-Peasy.AI.

## Features to look for

The details panel lists what each model can do:

| Feature | What it means |
| - | - |
| **Internet Access** | Can search the web. Needed for the **Search**, **Academic**, and **Deep Research** modes. |
| **Code Interpreter** | Can use Marky's computer: read your files, run code, create documents. Needed to attach files. |
| **Advanced Reasoning** | Thinks before answering. Better for hard problems, slower. |
| **Image Generation** | Can make and edit images in the chat. |

## All models

### OpenAI

| Model | Word rate | Free plan | Web search | Reads files |
| - | - | - | - | - |
| GPT-6 Astra | 2x | No | Yes | Yes |
| GPT-6 Astra (Max) | 2x | No | Yes | Yes |
| GPT-6 Sol | 0.5x | No | Yes | Yes |
| GPT-6 Sol (Max) | 0.5x | No | Yes | Yes |
| GPT-6 Luna | 0.1x | Yes | Yes | Yes |
| GPT-6 Luna (Max) | 0.1x | Yes | Yes | Yes |
| GPT-5.6 Sol | 1x | No | Yes | Yes |
| GPT-5.6 Sol (Max) | 1x | No | Yes | Yes |
| GPT-5.6 Terra | 0.5x | Yes | Yes | Yes |
| GPT-5.6 Terra (Max) | 0.5x | No | Yes | Yes |
| GPT-5.6 Luna | 0.2x | Yes | Yes | Yes |
| GPT-5.6 Luna (Max) | 0.2x | Yes | Yes | Yes |
| GPT-5.5 Instant | 1x | Yes | Yes | Yes |
| GPT-5.4 Instant | 0.5x | Yes | Yes | Yes |
| GPT-5.5 Thinking | 1x | No | Yes | Yes |
| GPT-5.5 Pro | 1x | No | Yes | Yes |
| GPT-5.4 Thinking | 0.5x | No | Yes | Yes |
| GPT-5.4 Pro | 0.5x | No | Yes | Yes |
| GPT-5.2 Instant | 0.5x | Yes | Yes | Yes |
| GPT-5.2 Thinking | 0.5x | No | Yes | Yes |
| GPT-5.2 Pro | 0.5x | No | Yes | Yes |
| GPT-5.1 Instant | 0.5x | Yes | Yes | Yes |
| GPT-5.1 Thinking | 0.5x | No | Yes | Yes |
| GPT-5.1 Pro | 0.5x | No | Yes | Yes |
| GPT-5 | 0.5x | No | Yes | Yes |
| GPT-5 mini | 0.25x | Yes | Yes | Yes |
| GPT-5 nano | Unlimited | Yes | Yes | Yes |
| GPT-5 Codex | 0.5x | No | Yes | Yes |
| GPT-5 Chat | 0.5x | No | Yes | Yes |
| GPT-4.1 | 0.5x | No | Yes | Yes |
| GPT-4.1 mini | 0.25x | Yes | Yes | Yes |
| GPT-4.1 nano | Unlimited | Yes | Yes | Yes |
| GPT-4o mini | Unlimited | Yes | Yes | Yes |
| GPT-4o | 0.5x | No | Yes | Yes |
| GPT-4 Turbo | 1x | No | Yes | Yes |
| GPT-3.5 Turbo | Unlimited | Yes | — | — |
| o3 | 1x | No | Yes | Yes |
| o4-mini | 0.25x | No | Yes | Yes |

### Anthropic

| Model | Word rate | Free plan | Web search | Reads files |
| - | - | - | - | - |
| Claude Fable 5.1 | 2x | No | Yes | Yes |
| Claude Fable 5.1 Reasoning (Low) | 2x | No | Yes | Yes |
| Claude Fable 5.1 Reasoning (Medium) | 2x | No | Yes | Yes |
| Claude Fable 5.1 Reasoning (High) | 2x | No | Yes | Yes |
| Claude Fable 5.1 Reasoning (Max) | 2x | No | Yes | Yes |
| Claude Fable 5 | 2x | No | Yes | Yes |
| Claude Fable 5 Reasoning (Low) | 2x | No | Yes | Yes |
| Claude Fable 5 Reasoning (Medium) | 2x | No | Yes | Yes |
| Claude Fable 5 Reasoning (High) | 2x | No | Yes | Yes |
| Claude Fable 5 Reasoning (Max) | 2x | No | Yes | Yes |
| Claude Opus 5.5 | 1x | No | Yes | Yes |
| Claude Opus 5.5 Reasoning (Medium) | 1x | No | Yes | Yes |
| Claude Opus 5.5 Reasoning (High) | 1x | No | Yes | Yes |
| Claude Opus 5.5 Reasoning (Max) | 1x | No | Yes | Yes |
| Claude Opus 5 | 1x | No | Yes | Yes |
| Claude Opus 5 Reasoning (Low) | 1x | No | Yes | Yes |
| Claude Opus 5 Reasoning (Medium) | 1x | No | Yes | Yes |
| Claude Opus 5 Reasoning (High) | 1x | No | Yes | Yes |
| Claude Opus 5 Reasoning (Max) | 1x | No | Yes | Yes |
| Claude Opus 4.8 | 1x | No | Yes | Yes |
| Claude Opus 4.8 Reasoning (Low) | 1x | No | Yes | Yes |
| Claude Opus 4.8 Reasoning (Medium) | 1x | No | Yes | Yes |
| Claude Opus 4.8 Reasoning (High) | 1x | No | Yes | Yes |
| Claude Opus 4.8 Reasoning (Max) | 1x | No | Yes | Yes |
| Claude Opus 4.7 | 1x | No | Yes | Yes |
| Claude Opus 4.7 Reasoning (Low) | 1x | No | Yes | Yes |
| Claude Opus 4.7 Reasoning (Medium) | 1x | No | Yes | Yes |
| Claude Opus 4.7 Reasoning (High) | 1x | No | Yes | Yes |
| Claude Opus 4.7 Reasoning (Max) | 1x | No | Yes | Yes |
| Claude Opus 4.6 | 1x | No | Yes | Yes |
| Claude Opus 4.6 Reasoning | 1x | No | Yes | Yes |
| Claude Sonnet 5 | 0.5x | No | Yes | Yes |
| Claude Sonnet 5 Reasoning | 0.5x | No | Yes | Yes |
| Claude Sonnet 5 Reasoning (Max) | 0.5x | No | Yes | Yes |
| Claude Sonnet 4.6 | 0.5x | No | Yes | Yes |
| Claude Sonnet 4.6 Reasoning | 0.5x | No | Yes | Yes |
| Claude Sonnet 4.5 Reasoning | 0.5x | No | Yes | Yes |
| Claude Opus 4.5 Reasoning | 1x | No | Yes | Yes |
| Claude Haiku 4.5 | 0.25x | No | Yes | Yes |
| Claude Haiku 4.5 Reasoning | 0.25x | No | Yes | Yes |
| Claude Sonnet 4.5 | 0.5x | No | Yes | Yes |
| Claude Opus 4.5 | 1x | No | Yes | Yes |

### Google

| Model | Word rate | Free plan | Web search | Reads files |
| - | - | - | - | - |
| Gemini 3.8 Flash | 0.25x | Yes | Yes | Yes |
| Gemini 3.7 Flash | 0.25x | Yes | Yes | Yes |
| Gemini 3.6 Flash | 0.25x | Yes | Yes | Yes |
| Gemini 3.5 Flash | 0.25x | Yes | Yes | Yes |
| Gemini 3 Flash | 0.25x | Yes | Yes | Yes |
| Gemini Flash 2.5 | Unlimited | Yes | Yes | Yes |
| Gemini 2.5 Pro | 0.5x | No | Yes | Yes |
| Gemini 3.1 Pro | 0.5x | Yes | Yes | Yes |

### xAI

| Model | Word rate | Free plan | Web search | Reads files |
| - | - | - | - | - |
| Grok 4.20 | 0.25x | No | Yes | Yes |
| Grok 4.20 Reasoning | 0.25x | No | Yes | Yes |
| Grok 4.5 | 0.25x | No | Yes | Yes |
| Grok 4.6 | 0.25x | No | Yes | Yes |
| Grok 4.3 | 0.25x | No | Yes | Yes |

### Meta

| Model | Word rate | Free plan | Web search | Reads files |
| - | - | - | - | - |
| Muse Spark 1.3 | 0.25x | Yes | Yes | Yes |
| Muse Spark 1.3 (Max) | 0.25x | Yes | Yes | Yes |

### Deepseek

| Model | Word rate | Free plan | Web search | Reads files |
| - | - | - | - | - |
| Deepseek V4 Pro | 0.25x | Yes | Yes | Yes |
| Deepseek V4.1 Flash | 0.05x | Yes | Yes | Yes |
| Deepseek V4.1 Flash (Max) | 0.05x | Yes | Yes | Yes |

### Perplexity

| Model | Word rate | Free plan | Web search | Reads files |
| - | - | - | - | - |
| Perplexity Sonar Pro | 0.5x | No | Yes | — |
| Perplexity Sonar | 0.25x | Yes | Yes | — |

### Moonshot

| Model | Word rate | Free plan | Web search | Reads files |
| - | - | - | - | - |
| Kimi K3 | 0.5x | Yes | Yes | Yes |
| Kimi K2.7 Code | 0.1x | Yes | Yes | Yes |
| Kimi K2.6 | 0.1x | Yes | Yes | Yes |

### Z.ai

| Model | Word rate | Free plan | Web search | Reads files |
| - | - | - | - | - |
| GLM 5.3 Flash (default) | 0.05x | Yes | Yes | Yes |
| GLM 5.3 Flash (Max) | 0.05x | Yes | Yes | Yes |
| GLM 5.3 (Max) | 0.25x | Yes | Yes | Yes |
| GLM 5.3 | 0.25x | Yes | Yes | Yes |

### MiniMax

| Model | Word rate | Free plan | Web search | Reads files |
| - | - | - | - | - |
| MiniMax M3 | 0.1x | Yes | Yes | Yes |

<AccordionGroup>
  <Accordion title="Why can't I pick a Grok or DeepSeek model?">
    While Gmail, Google Calendar, or Google Drive is connected and turned on, xAI (Grok) and DeepSeek models aren't available, because of Google's rules for how its data may be used. Turn the Google connectors off in [Connectors](/docs/marky/connectors) to use these models.
  </Accordion>

  <Accordion title="What do (Max) and Thinking mean in a model name?">
    They're the same model set to think longer before answering. They're better at hard problems, slower, and write more reasoning, so they use more words.
  </Accordion>

  <Accordion title="Why does Claude Fable ask me to confirm?">
    Claude Fable models count words at 2x, so the picker asks you to confirm when you choose one.
  </Accordion>
</AccordionGroup>


# Marky Agent
Source: https://easy-peasy.ai/docs/marky/overview

Marky is Easy-Peasy.AI's AI agent: it answers questions, researches the web, and does real work such as slide decks, websites, images, and scheduled reports.

Marky is an AI agent you talk to like a colleague. Ask a question and it answers. Give it a task and it gets the work done: it searches the web, reads your files, writes and runs code, uses your connected apps, and hands you finished results such as a slide deck, a website, a spreadsheet, or an image.

Open **Marky Agent** from the dashboard sidebar, or go to [easy-peasy.ai/chat](https://easy-peasy.ai/chat).

<Frame>
  <img alt="The Marky home screen: the question How can I help you today?, the message box with a + button, a microphone, and a send button, and the mode buttons Create slides, Build website, Images, Music, Video, and More, with More open showing Audio, Build app, Search, Academic, and Deep Research" />
</Frame>

## What Marky can do

<Columns>
  <Card title="Chat and get work done" icon="message" href="/docs/marky/chat">
    Ask anything, attach files, and follow Marky's work step by step.
  </Card>

  <Card title="Create slides" icon="presentation-screen" href="/docs/marky/slides">
    Get a finished deck from a topic or a document, as PowerPoint or PDF.
  </Card>

  <Card title="Build websites and apps" icon="globe" href="/docs/marky/websites">
    Describe a site or app, preview it, and publish it at your own address.
  </Card>

  <Card title="Images, music, and video" icon="photo-film" href="/docs/marky/media">
    Make pictures, songs, videos, and voiceovers without leaving the chat.
  </Card>

  <Card title="Research" icon="magnifying-glass" href="/docs/marky/research">
    Web search with sources, academic papers, and in-depth research reports.
  </Card>

  <Card title="Browse the web" icon="window-maximize" href="/docs/marky/browser">
    Let Marky use websites for you, in the cloud or in your own Chrome.
  </Card>

  <Card title="Connectors" icon="plug" href="/docs/marky/connectors">
    Connect Gmail, Google Calendar, Notion, GitHub, Stripe, and more.
  </Card>

  <Card title="Scheduled tasks" icon="calendar-clock" href="/docs/marky/scheduled-tasks">
    Have Marky do a job every day or week and send you the result.
  </Card>
</Columns>

## How Marky works

When you give Marky a task, it works in steps and shows each one in the chat: a web search, a command it runs, a file it writes. For longer jobs, it writes a short plan first, and a **Task progress** bar above the message box shows how far along it is.

Marky has its own computer in the cloud for each chat. It uses it to open websites, run code, and create files, and you can watch it work in the **Marky's Computer** panel. Files you attach are copied there too.

Marky keeps working even if you close the tab or switch to another chat. Come back later, and the result is waiting.

## Pick a mode, or just ask

The buttons under the message box of a new chat set Marky up for a kind of task:

| Mode | What it's for |
| - | - |
| **Create slides** | A presentation. Adds slide templates and deck options. See [Create slides](/docs/marky/slides). |
| **Build website** | A website or web app with a live preview. See [Build websites and apps](/docs/marky/websites). |
| **Images**, **Music**, **Video** | Example prompts for each. **Images** also shows the image model. See [Images, music, and video](/docs/marky/media). |
| **More** > **Audio** | Voiceovers and speech. |
| **More** > **Build app** | A mobile app you can preview on your phone. |
| **More** > **Search**, **Academic**, **Deep Research** | Answers from the web, from research papers, or from a multi-step research run. See [Research](/docs/marky/research). |

You don't need a mode. Marky picks the right tools from what you ask, so "make me a 5-slide deck about our Q3 results" works in a plain chat too.

## Make Marky your own

* [Choose a model](/docs/marky/models) for each chat, from fast and cheap to the most capable.
* Add [skills and prompts](/docs/marky/skills) for jobs you repeat, and a [brand voice](/docs/marky/prompts-and-brand-voice) for your writing style.
* Turn on [memory](/docs/marky/memory) so Marky remembers facts about you and your business across chats, or import your ChatGPT history.

## What it costs

Marky uses **words** from your plan for its replies and its work, at a rate that depends on the model. Images, slides, and videos it makes use image & video credits, and music and speech use text-to-speech characters. See [Models and words](/docs/marky/models) and [Plans and credits](/docs/account/plans-and-credits).


# Prompts and brand voice
Source: https://easy-peasy.ai/docs/marky/prompts-and-brand-voice

Reuse good prompts from the Prompt Library, save your own, and make Marky write in your brand's voice.

## Prompt Library

The Prompt Library has ready-made prompts for common jobs, such as social posts, ads, emails, job descriptions, and SEO briefs, plus the prompts you save yourself.

Open it with **Prompts** in the Marky sidebar, the sun icon at the top of a chat, or **+** > **Prompts** > **Open Prompt Library**.

<Frame>
  <img alt="The Prompt Library with a language picker, a search box, categories from All and My Prompts to Creative, and prompt cards such as Paragraph Generator, Explain like I am 5, Instagram Captions, and Facebook Post Ideas" />
</Frame>

1. Pick a category, or search for a prompt. The flag next to the title switches the language of the prompts.
2. Click **Use this prompt**. The prompt goes into the message box.
3. Replace the parts in square brackets, such as `[topic]` or `[product]`, with your details, and send.

### Save your own prompts

1. In the Prompt Library, open **My Prompts** and click **Add Prompt**.
2. Enter a **Title**, a **Category**, and the **Prompt**. To have the prompt written for you, click **Write for me**, describe the prompt you need, and click **Generate**.
3. Click **Save**.

Your ten newest prompts are also in **+** > **Prompts**, one click from any chat.

On a Teams plan, turn on **Share with Team** to give your teammates the prompt. **Hide the Prompt content from the team members** lets them use it without seeing the wording.

## Brand voice

A brand voice describes how your business writes: its tone, vocabulary, and style. When one is on, Marky writes in it.

Click **+** > **Brand Voice** and pick a voice. A megaphone button appears next to **+**; hover over it to see which voice is on, or click it to change it. The voice stays on, in this chat and new ones, until you choose **None** or reload the page.

To make one, click **+** > **Brand Voice** > **Create a brand voice**. You can build it from your website, writing samples, files, or a description of your style. See [Create a brand voice](/docs/brand-voice/create) and [Use a brand voice](/docs/brand-voice/use), which also covers the writing tools.

<Columns>
  <Card title="Skills" icon="bolt" href="/docs/marky/skills">
    Teach Marky a whole job, not just a prompt.
  </Card>

  <Card title="Memory" icon="brain" href="/docs/marky/memory">
    Let Marky remember facts about you and your business.
  </Card>
</Columns>


# Research
Source: https://easy-peasy.ai/docs/marky/research

Get answers from the web with sources, search academic papers, or have Marky write an in-depth research report.

Marky searches the web whenever a question needs current information, and it names the sites it used next to each point, so you can check them. For research that needs more than a quick search, use one of the research modes.

## Research modes

Start a new chat, click **More** under the message box, and pick a mode. Then ask your question as usual.

| Mode | What Marky does | Good for |
| - | - | - |
| **Search** | Searches the web and answers with sources. | Current facts, prices, news, comparisons. |
| **Academic** | Searches research papers only. | Literature reviews, citations, scientific questions. |
| **Deep Research** | Runs many searches in a row, reads the pages, follows leads, and writes a structured report. | Market research, competitor analysis, "everything about X" questions. Takes several minutes. |

All three need a model with **Internet Access**. If the current model doesn't have it, you'll see a message that the model doesn't support the mode; pick another model. See [Models and words](/docs/marky/models).

<Frame>
  <img alt="Marky's answer on the left with a source after each point, and the Marky's Computer panel on the right listing the 15 search results it used" />
</Frame>

## Check the sources

* Each source appears as a small link after the sentence it supports. Click it to open the page.
* Open **Marky's Computer** to see every search Marky ran and all the results it looked at.

## Read websites that need a browser

For pages that search can't read, Marky opens them in a real browser. Reddit, LinkedIn, and Facebook always go through the browser. See [Browse the web](/docs/marky/browser).

## Tips

* Say how fresh the information must be: "only sources from 2026".
* Ask for the format you need: "a table comparing price, features, and reviews", "a one-page summary with citations".
* Follow up in the same chat to dig deeper into one point. Marky keeps the earlier research in mind.
* For research you repeat, such as a weekly competitor check, set up a [scheduled task](/docs/marky/scheduled-tasks).

## Credits

Research uses words at the rate of the chat's model, for both the searching and the answer. **Deep Research** reads many pages, so it uses more words than a normal question.


# Scheduled tasks
Source: https://easy-peasy.ai/docs/marky/scheduled-tasks

Have Marky do a task on a schedule, such as a Monday morning report, and get the result in the chat and by email.

A scheduled task is a job Marky does on its own at set times: every weekday at 8, every Monday at 9, or once next Friday. At each run, Marky follows the task's instructions with all its usual tools, such as web search, your connected apps, and file creation, posts the result in the chat, and emails you a link.

Use scheduled tasks for work that repeats: a weekly content plan, a daily news digest on your industry, a Monday check of your inbox or analytics, or a reminder-style report before a meeting.

## Create a task by asking

The easiest way is to tell Marky what you want and when, in any chat:

> Every Monday at 9am, write three Instagram post ideas for Sunrise Yoga, a small yoga studio in Austin, and post them here.

Marky writes detailed instructions for the task, sets the schedule, and confirms when the first run will be. The task runs in the same chat.

<Frame>
  <img alt="The Scheduled Tasks window with one active task, sunrise-yoga-monday-ig-ideas, set to Weekly on Monday at 09:00 with the next run on September 28" />
</Frame>

Say the time the way you'd say it to a person: "every weekday at 8:30", "on the first of each month", "every 2 hours during work hours", "once, next Friday at 5pm". Tasks run in your time zone.

## Create or change a task yourself

Click **Scheduled** in the Marky sidebar to see all your tasks.

* **New task** creates one from scratch.
* The pause button stops a task without deleting it. Click it again to resume.
* The **⋯** menu has **Edit**, **View chat** (opens the chat where the results go), and **Delete**.

<Frame>
  <img alt="The Edit scheduled task window with a title, a Weekly on Monday at 09:00 schedule in the Asia/Singapore time zone, a Set expiration date checkbox, the prompt, and Delete, Cancel, and Save buttons" />
</Frame>

| Field | What it does |
| - | - |
| **Title** | The name shown in the list and in emails. |
| **Schedule** | **Daily**, **Weekly**, **Monthly** (on a day from 1 to 28), or **No Repeat** for a single run, plus the time. The time zone under it is the one the task runs in. |
| **Prompt** | What Marky should do on each run. Describe the job, not the timing. |

When you edit a task, you also get:

* **Set expiration date**, which stops the task after a date, for example at the end of a campaign.
* **Custom**, for schedules the simple options can't show, such as "weekdays at 8:30" that Marky set up. It's a cron expression with seconds first: `0 30 8 * * 1-5` means 8:30 on weekdays. Tasks that repeat every few minutes or hours show **Every N minutes** instead.

Each task shows its status: **Active**, **Running**, **Paused**, **Completed** (a one-time task that ran), **Failed**, or **Expired**.

## Write good task instructions

At run time you're not there to answer questions, so the prompt has to stand on its own. When Marky creates a task for you, it writes the prompt this way. When you write one yourself, include:

* **The goal and who it's for**: "Write three Instagram post ideas for Sunrise Yoga, a small yoga studio in Austin."
* **Where to look**: the websites, files, or connected apps to use.
* **How to avoid repeats**: "Check the earlier runs in this chat and don't repeat an idea."
* **The format of the result**: a table, three bullets, a one-page report.

## Limits and costs

* You can have up to 10 active tasks. Pause or delete one to add another.
* A task can run at most every 5 minutes. Daily or weekly schedules suit most work.
* Each run uses words like any other message to Marky, plus credits for anything it generates, such as images. Frequent schedules add up, so check [Usage](/docs/account/usage) after the first few runs.

## Results and emails

* The result of every run appears in the task's chat. A task you create by asking runs in that chat. A task you create with **New task** gets a chat of its own, named after the task. Open it with **View chat** or from the email.
* After each run, you get an email that the task ran, with a link to the result.
* If a run fails, for example because you're out of words, you get one email about it. You won't get another until the task runs successfully again.

<AccordionGroup>
  <Accordion title="Can a task use my Gmail, Calendar, or other apps?">
    Yes. A scheduled run can use every app you've connected in **Connectors**, so a task like "every morning, summarize the unread emails from clients" works. See [Connectors](/docs/marky/connectors).
  </Accordion>

  <Accordion title="A task ran at the wrong time">
    Tasks run in the time zone shown under **Schedule** when you edit the task. If it's wrong, edit the task and save it again, or tell Marky the time zone you want.
  </Accordion>

  <Accordion title="Can Marky ask me questions during a run?">
    No. Scheduled runs happen without you, so Marky can't ask for approval or clarification. If something needs your input, it says so in the result.
  </Accordion>
</AccordionGroup>

<Columns>
  <Card title="Connectors" icon="plug" href="/docs/marky/connectors">
    Let Marky work with Gmail, Google Calendar, Notion, and more.
  </Card>

  <Card title="Usage" icon="chart-simple" href="/docs/account/usage">
    See how many words your tasks use.
  </Card>
</Columns>


# Skills
Source: https://easy-peasy.ai/docs/marky/skills

Teach Marky how you want a job done once, as a skill, and it follows those instructions every time the job comes up.

A skill is a set of instructions, and optionally files such as templates or scripts, that teaches Marky how to do a specific job your way: "write our weekly newsletter", "review a contract against our checklist", "turn sales notes into a CRM update".

Marky sees the name and description of each active skill. When a request matches one, Marky reads the skill first and follows it. You can also ask for a skill by name.

## Manage skills

Click **Skills** in the Marky sidebar.

<Frame>
  <img alt="The Manage Skills window with the Official skills skill-creator and automation-and-scheduling, each with an on/off switch and a menu" />
</Frame>

* The switch on each card turns a skill on or off. Marky only uses skills that are on.
* **Official** skills come from us. **Custom** skills are yours.
* The **⋯** menu has **Try it**, **Preview**, **Download**, and **Delete**.

Two official skills are included: **skill-creator**, which helps Marky write new skills with you, and **automation-and-scheduling**, which helps it set up [scheduled tasks](/docs/marky/scheduled-tasks) and automations.

## Add a skill

Click **Add** and choose how:

<Frame>
  <img alt="The Add menu with Build with Marky, Write skill instructions, Upload a skill, and Import from GitHub" />
</Frame>

| Option | How it works |
| - | - |
| **Build with Marky** | Opens a chat where Marky interviews you about the job and writes the skill. When it's done, click **Add to My Skills**. The easiest way to start. |
| **Write skill instructions** | Type a name, a short description of when to use it, and the instructions yourself. |
| **Upload a skill** | Upload a `.zip` or `.skill` file, up to 50 MB, with a `SKILL.md` file at the top level. |
| **Import from GitHub** | Paste the link to a public GitHub repository that contains a skill. |

Skills use the same format as Claude's Agent Skills, so skills you've written for Claude work in Marky too.

## Use a skill

* **Let Marky decide**: ask as usual. If your request matches an active skill's description, Marky uses it.
* **Pick it yourself**: click **+** > **Skills** and choose one, or type `/` followed by the skill's name in your message, such as `/weekly-newsletter`.

## Write a good skill

* **Description**: say exactly when the skill applies: "Use when the user asks for the weekly customer newsletter." Marky decides from this line alone.
* **Instructions**: write the steps the way you'd brief a new colleague: the sources to use, the structure of the result, the tone, what to avoid, and how to check the work.
* **Files**: add templates, examples of good results, or reference documents to the skill's folder, and mention them in the instructions.

Test a new skill with **Try it**, then refine it.

<Columns>
  <Card title="Prompts and brand voice" icon="book-open" href="/docs/marky/prompts-and-brand-voice">
    Save prompts you reuse, and set the tone for Marky's writing.
  </Card>

  <Card title="Scheduled tasks" icon="calendar-clock" href="/docs/marky/scheduled-tasks">
    Run a skill on a schedule.
  </Card>
</Columns>


# Create slides
Source: https://easy-peasy.ai/docs/marky/slides

Ask Marky for a presentation, approve the style on the title slide, and download the deck as PowerPoint or PDF.

Marky can turn a topic, notes, or a document into a finished slide deck. You describe what the deck is about, Marky writes the outline, designs the slides, and shows them in a panel next to the chat, where you can present them or download them.

<Frame>
  <img alt="A finished four-slide deck about a yoga studio: the chat on the left shows a delivery card with 4 slides and 9 credits used, and the panel on the right shows the Who We Are slide with the Present and Save buttons" />
</Frame>

## Make a deck

<Steps>
  <Step title="Switch to Slides">
    Start a new chat and click **Create slides** under the message box. A **Slides** chip appears in the box, with the options for the deck next to it. To leave Slides mode, hover over the chip and click the **✕**.

    <Frame>
      <img alt="The message box in Slides mode with the Slides, Creative, engine, and 4K buttons, and the Choose a template gallery below it with Upload your deck, Startup Pitch, Product Launch, Creative Bold, Corporate Modern, Minimal Elegant, and more" />
    </Frame>
  </Step>

  <Step title="Pick a template (optional)">
    Under **Choose a template**, click a style such as **Startup Pitch**, **Annual Report**, or **Tech Dark**. The eye icon on a card opens a preview with **Use this template**. If you don't pick one, Marky designs a look that fits your topic.

    Most templates make Creative slides, and a few make HTML slides. Picking one turns **Creative** on or off to match. To use your own brand look, click **Upload your deck**. See [Use your own deck as a template](#use-your-own-deck-as-a-template).
  </Step>

  <Step title="Describe the deck">
    Say what the deck is about, who it's for, and roughly how many slides you want. For example: "A 4-slide deck introducing Sunrise Yoga, a small yoga studio in Austin: who we are, our classes, pricing, and how to book a first class." You can also attach notes, a report, or a spreadsheet with **+** and ask Marky to build the deck from it.
  </Step>

  <Step title="Approve the title slide">
    With **Creative** on, Marky makes the title slide first and asks whether you like the style. Choose **Yes, continue** to make the rest, or **Adjust style first** and say what to change, such as the colors, the mood, or the imagery. This way you only pay for the whole deck once you like the look.

    <Frame>
      <img alt="The title slide of the yoga studio deck in the panel on the right, and a Style check question in the chat asking whether to continue, with the options Yes, continue, Adjust style first, and Other" />
    </Frame>
  </Step>

  <Step title="Present or download">
    The finished deck opens in the panel on the right. Click a slide number to look at a slide, **Present** to show the deck full screen, or **Save** > **Save as PPTX** or **Save as PDF** to download it. **Save** works once every slide is finished. To open the panel again later, click **Open** on the deck card in the chat.
  </Step>
</Steps>

## Creative and HTML slides

The **Creative** button in the message box decides how the slides are made. It's on by default.

| | **Creative** on | **Creative** off (HTML slides) |
| - | - | - |
| How slides are made | Each slide is an image, drawn by an AI image model | Each slide is a web page with real text, shapes, and pictures |
| Best for | Pitch decks, keynotes, and anything where the look matters most | Reports, lessons, and decks you want to keep editing |
| Title slide first | Yes, so you can approve the style | No, Marky makes the whole deck |
| Editing afterwards | Ask Marky to change a slide; it redraws it | Click **Edit** in the panel and type over the text, ask Marky, or edit in PowerPoint |
| PowerPoint download | Each slide is a picture | Text, shapes, and pictures stay editable |
| Credits | Image & video credits for every slide (see below) | Words for Marky's work; pictures are free stock photos when a good one exists, otherwise an AI image that uses credits |

A Creative deck can have up to 15 slides.

### Image engine and resolution

With **Creative** on, two more buttons appear:

* **Engine**: the OpenAI logo is **GPT Image 2.5 (OpenAI)**, the default. The Google logo is **Nano Banana 2 (Google)**.
* **4K**: makes each slide in 4K instead of the standard 2K. It's sharper, for printing or very large screens, and costs more.

| Engine | 2K | 4K |
| - | - | - |
| GPT Image 2.5 (OpenAI) | 3 credits per slide | 11 credits per slide |
| Nano Banana 2 (Google) | 6 credits per slide | 8 credits per slide |

The title slide is charged when Marky makes it; the other slides are charged when you approve. A 4-slide deck at 2K with GPT Image 2.5 costs 12 credits: 3 for the title slide and 9 for the rest. The deck card in the chat shows only the second part.

## Change a deck

Ask in the same chat. For example:

* "Make slide 3 about our weekend workshops instead."
* "On slide 2, change 'Founded 2018' to 'Founded 2019'."
* "Use a darker background on every slide."

Marky changes only the slides you mention. With **Creative** on, each redrawn slide costs the same as a new one. In HTML decks you can also add a slide in the middle, for example "Add a slide with our teachers after slide 2", and the rest of the deck stays as it is.

## Use your own deck as a template

On a paid plan, you can upload a deck in your brand's style, and Marky makes new slides that look like it.

1. In Slides mode, click **Upload your deck** under **Choose a template**.
2. Drop a PowerPoint (PPTX), a PDF, or images of your slides. Files can be up to 50 MB.
3. Click **Save template**. Marky studies the colors, fonts, and layouts, and adds the template to your gallery.
4. Pick your template, then describe the new deck.

<AccordionGroup>
  <Accordion title="Can I make slides on the Free plan?">
    Yes. HTML slides (**Creative** off) work on every plan and use words. For a Creative deck, the Free plan makes only the title slide so you can see the style; the full Creative deck and uploading your own templates need a paid plan.
  </Accordion>

  <Accordion title="The text on a Creative slide has a typo">
    Creative slides are images, so you can't retype them. Ask Marky to fix the slide, for example: "On slide 2, change 'Founded 2018' to 'Founded 2019'." For decks you'll edit a lot, turn **Creative** off and use HTML slides.
  </Accordion>

  <Accordion title="Where do I find a deck later?">
    Decks stay in the chat where you made them. Open the chat from the sidebar and click **Open** on the deck card. Download the files if you want to keep a copy: HTML slides live in Marky's workspace, which is cleared some time after you stop using the chat.
  </Accordion>

  <Accordion title="What happens if I run out of credits halfway?">
    Marky stops at the last slide your credits cover and tells you. The slides made so far stay in the panel, and you can download them. Buy more credits or wait for your plan to renew, then ask Marky to finish the deck.
  </Accordion>
</AccordionGroup>

<Columns>
  <Card title="Build a website" icon="globe" href="/docs/marky/websites">
    Turn a description into a live website or web app.
  </Card>

  <Card title="Plans and credits" icon="coins" href="/docs/account/plans-and-credits">
    How words and image & video credits work.
  </Card>
</Columns>


# Build websites and apps
Source: https://easy-peasy.ai/docs/marky/websites

Describe a website or web app, watch Marky build it, preview it, and publish it at your own address.

Marky can build a complete website or web app from a description: a landing page for a launch, a portfolio, a booking site, a dashboard, a small game, or an app with sign-in and a database. You see the result in a live preview as it's built, ask for changes in plain language, and publish it when you're happy.

<Frame>
  <img alt="A finished website build: in the chat, a Website delivered card with a thumbnail and an Open button, and Task progress at 3 of 3; on the right, the workspace's Browser tab showing the yoga studio site" />
</Frame>

## Build a site

<Steps>
  <Step title="Switch to Build website">
    Start a new chat and click **Build website** under the message box.

    <Frame>
      <img alt="The message box in Website mode, a Recommendation to use Claude Opus 5 with Dismiss and Switch model buttons, and Browse templates with the categories Landing page, Dashboard, Portfolio, Corporate, SaaS, Link-in-bio, and E-commerce, and ideas such as Product launch page and Event registration page" />
    </Frame>

    Marky suggests **Claude Opus 5** for building, because it makes fewer mistakes on large projects. Click **Switch model** to use it, or **Dismiss** to stay on your current model. See [Models and words](/docs/marky/models) for what each model costs.
  </Step>

  <Step title="Describe the site">
    Say what the site is for, the sections you want, and the look. For example: "A one-page website for Sunrise Yoga, a small yoga studio in Austin: a hero section, our classes, a weekly schedule, pricing, and a contact form."

    For ideas, pick a category under **Browse templates** and click one. To copy the look of a site you like, click **Use reference URL** and paste its address; Marky takes a screenshot and uses it as a visual reference.
  </Step>

  <Step title="Watch Marky build">
    Marky plans the work, finds images, writes the code, and checks the result in its own browser. The workspace opens on the right. In the **Code** tab you can watch each file being written.

    <Frame>
      <img alt="Marky writing the site's files in the chat, with Writing file Classes.tsx and Schedule.tsx steps, and the Code tab on the right showing the project's files and the schedule component being written" />
    </Frame>
  </Step>

  <Step title="Preview and refine">
    When the site is ready, Marky posts a **Website delivered** card; click **Open** to see it again later. The **Browser** tab shows the live preview. Switch between the desktop and mobile view, refresh it, or open it in a new tab. Then ask for changes in the chat: "Make the hero image brighter", "Add a FAQ section above the contact form", "Use our logo from the attached file."

    <Frame>
      <img alt="The Browser tab of the workspace showing the finished yoga studio site: a navigation bar, a Book a class button, the headline Begin your day the way the sun does, slowly, and with intention, and buttons for the weekly schedule and pricing" />
    </Frame>
  </Step>

  <Step title="Publish">
    Click **Publish** at the top of the workspace. Building and publishing takes a few minutes. When it's done, **Published Successfully!** shows your site's address, which ends in `.easy-peasy.site` and is based on the project's name. Copy it, share it on social media, or click **Visit Your Site**.

    <Frame>
      <img alt="The Published Successfully! window with the site URL, a Copy button, share buttons for LinkedIn, Facebook, X.com, and Reddit, and a Visit Your Site button" />
    </Frame>

    You can also ask Marky to publish the site; it asks you to confirm first.
  </Step>
</Steps>

<Note>
  Changes you make after publishing show up in the preview right away, but not on your live site. Click **Publish** again, or **Re-publish** in **Settings** > **General**, to put them live.
</Note>

## Websites, web apps, and mobile apps

Marky picks the right setup from your description:

| Project | What you get |
| - | - |
| **Website** | Pages, sections, forms, and images. Most landing pages, portfolios, and company sites. |
| **Web app** | Everything in a website, plus user accounts, a database, and AI features. Ask for it when people need to sign in or save data: "a booking app where customers sign up and reserve a class". |
| **Mobile app** | An app built with Expo. Start it with **More** > **Build app** on a new chat. See [Mobile apps](#mobile-apps). |

## The workspace

| Tab | What it's for |
| - | - |
| **Browser** | The live preview, with **Desktop view** and **Mobile view**, **Refresh**, **Open in new tab**, and **Copy URL**. Mobile apps show **Preview** instead. |
| **Code** | The project's files. Open any file to read it. |
| **Database** | The app's tables and their rows, for web apps with a database. |
| **Storage** | Files your app's visitors upload, such as profile pictures. Copy their links or delete them. |
| **Settings** | Publishing, domains, database access, secrets, and GitHub. See below. |

The **⋯** menu has **Version history**, **Push to GitHub**, and **Download ZIP**. To show or hide the workspace, use the monitor icon at the top of the chat.

### If the preview says the sandbox is paused

Marky's computer pauses when a project isn't used for a while. Click **Resume Sandbox** to wake it up; your files are kept. Published sites aren't affected: they keep running on their own.

## Mobile apps

For a mobile app, the **Preview** tab shows the app in the browser and a QR code. Scan it with the Expo Go app on your phone to try the app on a real device.

When the app is ready, click **Build** at the top of the workspace:

* **Android**: choose **Build APK** for a file you can install on any Android phone for testing, or **Build for Google Play** for an AAB file to upload to the Play Store. A build takes 5 to 15 minutes, then a download button appears. Android builds need a paid plan.
* **iOS**: coming soon.

## Undo changes with version history

Marky saves versions of your project as it works. Open **⋯** > **Version history** to see them, newest first.

<Frame>
  <img alt="Version history with three versions: the newest marked Live with a Visit button, and two older versions with Rollback buttons, each with a description, date, and thumbnail" />
</Frame>

* The published version is marked **Live**. **Visit** opens it.
* If your newest version isn't published yet, it's marked **Current** and has a **Publish** button.
* **Rollback** returns the project to an older version. You're asked whether to roll back the live site too.

## Settings

Click the gear icon in the workspace.

<Frame>
  <img alt="General Settings for a published site: the project name, the published URL with copy and open buttons, a Re-publish button, and an Always-On Hosting switch with its price" />
</Frame>

| Section | What's in it |
| - | - |
| **General** | The published address, **Re-publish** to put your latest changes live, and **Always-On Hosting**. |
| **Domains** | Your easy-peasy.site address and custom domains. |
| **Database** | The database address for web apps, with **Download as .env**. |
| **Secrets** | API keys and other private settings your app needs, such as a payment provider's key. Click **Add Secret**, then **Save**. They stay on the server and never appear in the site's code. |
| **GitHub** | Connect a GitHub account to keep a copy of the code in a repository. |

### Change the address

Your site gets a free address ending in `easy-peasy.site`. To change the first part, open **Settings** > **Domains**, click **Edit** next to **Easy-Peasy subdomain**, type the new name, and click **Save subdomain**. The new address works right away, and the old one stops working, so update any links you've shared.

<Frame>
  <img alt="The Domains settings with the Easy-Peasy subdomain being edited to example-yoga-demo, Cancel and Save subdomain buttons, and below them the Custom Domains options Buy a domain and Connect existing domain" />
</Frame>

### Use your own domain

Publish the site first, then open **Settings** > **Domains** > **Custom Domains**:

* **Buy a domain**: search for a name, check the yearly price, and click **Buy**. You pay by card, and the domain is connected for you. Domains are registered for one year.
* **Connect existing domain**: enter a domain you own and click **Add Domain**, then add the DNS record shown at your domain registrar. The status goes from **Pending DNS** to **SSL Provisioning** to **Active**. Click **Check again** to refresh it.

If you connect a main domain such as `example.com`, `www.example.com` works too.

### Always-On Hosting

Published sites go to sleep when nobody visits them, and the first visit afterwards takes a few seconds to load. **Always-On Hosting** keeps the site awake, so every visit is fast. It's a paid add-on for each site, available on paid plans. You'll see the price before you add it in **Settings** > **General**.

## Keep a copy of the code

* **Download ZIP**: **⋯** > **Download ZIP** gives you the whole project.
* **Push to GitHub**: **⋯** > **Push to GitHub** creates a new repository (private or public) or links an existing one, then pushes the code. Use **Sync Changes** later to push new versions.

For safety, Marky never pushes your secrets or `.env` files to GitHub.

## Credits

Building uses words, like any Marky task, at the rate of the model you pick. Images Marky generates for the site use image & video credits; stock photos are free. Publishing is free. Custom domains and **Always-On Hosting** are paid separately.

<AccordionGroup>
  <Accordion title="Is my published site public?">
    Yes. Anyone with the address can visit it. The preview in the **Browser** tab uses a private, hard-to-guess address, but it isn't password-protected either.
  </Accordion>

  <Accordion title="What happens to my site if I cancel my plan?">
    If you built the site on a paid plan and the plan ends, the site's hosting is paused. Renew your plan to bring it back.
  </Accordion>

  <Accordion title="How do I take a site offline?">
    Open **Settings** > **Data Controls** on the dashboard, find the site under **My Websites** or **My Apps**, and delete it. Deleting the chat where you built it also takes the site offline. See [Manage your chats](/docs/marky/manage-chats#delete-a-chat).
  </Accordion>

  <Accordion title="Can I edit the code myself?">
    Yes. Download the ZIP or push it to GitHub and work on it with your own tools. To keep building with Marky, ask for changes in the chat.
  </Accordion>
</AccordionGroup>

<Columns>
  <Card title="Create slides" icon="presentation-screen" href="/docs/marky/slides">
    Turn a topic or a document into a finished deck.
  </Card>

  <Card title="Models and words" icon="microchip" href="/docs/marky/models">
    Pick the right model for building.
  </Card>
</Columns>


# MCP Server
Source: https://easy-peasy.ai/docs/mcp/introduction

Connect Claude, Cursor, Windsurf, and other AI agents to Easy-Peasy.AI with the Model Context Protocol

## Overview

Easy-Peasy.AI runs a remote [MCP (Model Context Protocol)](https://modelcontextprotocol.io) server, so AI agents can use the platform directly — generate text from templates, create images and videos, synthesize speech, and transcribe audio.

Any MCP-capable client works: Claude, Claude Code, Cursor, Windsurf, ChatGPT connectors, or your own agent.

<Columns>
  <Card title="Endpoint" icon="link">
    `https://easy-peasy.ai/mcp`
  </Card>

  <Card title="Transport" icon="bolt">
    Streamable HTTP (stateless)
  </Card>
</Columns>

<Note>
  All usage is billed to your account exactly like the [REST API](/docs/api-reference/introduction) — same credits, word quotas, and plan limits.
</Note>

## Authentication

You can connect in one of two ways.

<AccordionGroup>
  <Accordion title="OAuth sign-in (recommended)" icon="right-to-bracket">
    The server implements the MCP authorization spec, so clients like claude.ai and Claude Desktop discover the flow automatically. Paste the URL, click **Connect**, sign in to Easy-Peasy.AI, and approve. No key handling.
  </Accordion>

  <Accordion title="API key" icon="key">
    Send your API key — the same one used by the REST API — as an `x-api-key` header or as `Authorization: Bearer YOUR_API_KEY`. Use this for clients that read headers from a config file.

    <Card title="Get your API key" icon="key" href="https://easy-peasy.ai/settings/api">
      Generate an API key from your Easy-Peasy.AI settings page.
    </Card>
  </Accordion>
</AccordionGroup>

## Connect your client

### Claude (claude.ai and Claude Desktop)

1. Go to **Settings → Connectors → Add custom connector**.
2. Paste `https://easy-peasy.ai/mcp` and click **Add**.
3. Click **Connect**, sign in to Easy-Peasy.AI, and approve access.

### Claude Code

```bash theme={null}
claude mcp add easy-peasy --transport http https://easy-peasy.ai/mcp \
  --header "x-api-key: YOUR_API_KEY"
```

### Cursor, Windsurf, and generic `mcp.json`

```json theme={null}
{
  "mcpServers": {
    "easy-peasy": {
      "url": "https://easy-peasy.ai/mcp",
      "headers": {
        "x-api-key": "YOUR_API_KEY"
      }
    }
  }
}
```

### Claude Desktop config file

Claude Desktop connects best through **Settings → Connectors** (OAuth, above). If you prefer the config file with an API key, bridge the connection with [`mcp-remote`](https://www.npmjs.com/package/mcp-remote):

```json theme={null}
{
  "mcpServers": {
    "easy-peasy": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "https://easy-peasy.ai/mcp",
        "--header",
        "x-api-key: YOUR_API_KEY"
      ]
    }
  }
}
```

## Verify the connection

Ask your agent to call the `get_account` tool. It returns `{ "connected": true }` to confirm authentication without spending generation credits. It does not expose account details or check your remaining quota.

<Card title="Browse the tools" icon="wrench" href="/docs/mcp/tools">
  See all 13 tools, their parameters, and the async polling pattern.
</Card>

## Inline media previews

Ask your agent to show a saved image or video with `display_media`. Clients that support MCP Apps can show an image gallery or a playable video inside the conversation. Other clients receive the result links.

Viewing an existing result does not consume generation credits. The video viewer can check a processing job without creating another video.

## Troubleshooting

<AccordionGroup>
  <Accordion title="I see links but no image or video viewer">
    Refresh the connector's tool list and ask your agent to use `display_media` with an existing image or video ID. In Claude, open **Customize → Connectors → Yours → Easy-Peasy.AI**, then choose **Refresh tools list** from the menu. Your client must support MCP Apps to render the viewer; links remain available otherwise.
  </Accordion>

  <Accordion title="401 Unauthorized">
    The API key is missing or invalid, or the OAuth token expired. Check the header name (`x-api-key`), or reconnect the connector to refresh the OAuth token.
  </Accordion>

  <Accordion title="A tool returns a plan error">
    `generate_image` and `transcribe_audio` require a paid plan for API access. Free accounts can still use `generate_text`, `generate_speech`, and the read-only tools.
  </Accordion>

  <Accordion title="Video or transcription never finishes">
    Both are asynchronous. The agent must poll `get_video` or `get_transcription` until the result is ready — see [Tools](/docs/mcp/tools).
  </Accordion>
</AccordionGroup>


# Tools
Source: https://easy-peasy.ai/docs/mcp/tools

The 13 tools exposed by the Easy-Peasy.AI MCP server

Generation tools use the documented [REST API](/docs/api-reference/introduction) and the same account credits and quotas. Viewing saved media and checking video status do not spend generation credits.

| Tool | Mode | Description |
| - | - | - |
| `get_account` | sync | Verify an explicitly requested connection check; returns only `connected: true`. |
| `list_templates` | sync | List text templates (presets) with slugs and categories. |
| `get_template` | sync | Input fields of one template (`keywords`, `extra1`…`extra14`). |
| `generate_text` | sync | Run a template; returns the generated text. |
| `generate_image` | sync | Generate or edit an image; returns URL(s). Paid plans only. |
| `get_image` | sync | Fetch a generated image record by id. |
| `generate_video` | **async** | Start a video generation; returns an `id`. |
| `get_video` | sync | Poll a video by `video_id` until `status` is `completed`. |
| `display_media` | sync | Show existing images or videos in an embedded gallery or player. |
| `list_voices` | sync | Text-to-speech voices, including your custom and cloned voices. |
| `generate_speech` | sync | Text-to-speech; returns the audio URL. |
| `transcribe_audio` | **async** | Start transcription of an audio or video URL; returns a `uuid`. Paid plans only. |
| `get_transcription` | sync | Poll a transcription by `audio_id`. |

Expand a tool below for its parameters. Parameters marked with `*` are required.

## Text generation

<AccordionGroup>
  <Accordion title="list_templates" icon="list">
    Lists a page of templates plus the list of categories. Returns `total`, `offset`, and `next_offset` (`null` on the final page).

    | Parameter | Type | Description |
    | - | - | - |
    | `category` | string | Optional category name to filter by. |
    | `limit` | integer | Maximum results, default `20`, maximum `50`. |
    | `offset` | integer | Start position, default `0`. Use the previous page's `next_offset`. |
  </Accordion>

  <Accordion title="get_template" icon="circle-info">
    Returns one template with its input fields (name, label, type, options). Field names such as `keywords` and `extra1`…`extra14` map directly onto `generate_text` parameters.

    | Parameter | Type | Description |
    | - | - | - |
    | `slug` \* | string | Template slug from `list_templates`. |
  </Accordion>

  <Accordion title="generate_text" icon="pen">
    Runs a template and returns the generated text. Uses the account word quota and API token allowance.

    | Parameter | Type | Description |
    | - | - | - |
    | `preset` \* | string | Template slug, for example `blog-post-intro`. |
    | `keywords` | string | Main topic or input text. Required by almost every template. |
    | `tone` | string | Tone of voice, for example `Friendly`. |
    | `language` | string | Output language name in English, for example `Spanish`. |
    | `outputs` | number | How many variations to generate (default `1`). |
    | `extra_fields` | object | Template-specific fields keyed by field name, for example `{"extra1": "value"}`. |
  </Accordion>
</AccordionGroup>

<Tip>
  Call `get_template` before `generate_text` so the agent knows which `extra_fields` a template expects.
</Tip>

## Images

<AccordionGroup>
  <Accordion title="generate_image" icon="image">
    Generates or edits an image and waits for the result — this can take a few minutes. Uses image credits and requires a paid plan.

    | Parameter | Type | Description |
    | - | - | - |
    | `prompt` \* | string | Description of the image. |
    | `model` | string | Model name, for example `Nano Banana 2`, `FLUX.2 [max]`, `Seedream 5.0 Pro`. |
    | `dimensions` | string | Size as `WIDTHxHEIGHT`, for example `1024x1024` (default). |
    | `outputs` | number | Number of images (default `1`). |
    | `style` | string | Optional style preset name. |
    | `resolution` | string | `1K` or `2K`, for models that support it. |
    | `source_images` | string\[] | Up to 10 image URLs to edit or transform (image-to-image). |
  </Accordion>

  <Accordion title="get_image" icon="magnifying-glass">
    | Parameter | Type | Description |
    | - | - | - |
    | `image_id` \* | number | Numeric image id returned by `generate_image`. |
  </Accordion>
</AccordionGroup>

## Video

<AccordionGroup>
  <Accordion title="generate_video" icon="video">
    Starts a generation and returns an id immediately. Defaults to MiniMax H3 Max, the same video model used by Marky Agent: 5 seconds at 768p with native audio, billed at 4 credits per second at 768p. With a first-frame image and no model, uses MiniMax H3 Max Image.

    | Parameter | Type | Description |
    | - | - | - |
    | `prompt` \* | string | Description of the video. |
    | `model` | string | Omit for `MiniMax H3 Max`, or `MiniMax H3 Max Image` when `image` is supplied. You can explicitly select another model, for example `Veo 3.1`, `Veo 3.1 Fast`, or `Seedance 2.5`. |
    | `duration` | string | MiniMax H3 Max accepts `"5"` through `"15"` (default `"5"`). Veo 3.1 text models accept `"4"`, `"6"`, or `"8"`; their Image variants accept only `"8"`. A trailing `s` is also accepted. If omitted, uses a supported model default. |
    | `aspect_ratio` | string | For example `16:9` (default), `9:16`, `1:1`. |
    | `resolution` | string | MiniMax H3 Max supports `768p` (default) or `480p`. Other models may support `720p`, `1080p`, or `2k`. |
    | `image` | string | First-frame URL. With no `model`, selects MiniMax H3 Max Image. When you supply a model, choose its Image variant, such as `Veo 3.1 Fast Image`. |
    | `generate_audio` | boolean | Controls sound on models with an audio toggle. MiniMax H3 Max always generates audio; `false` returns an error before generation. For a silent video, explicitly select a model with an audio toggle, such as Veo 3.1 Fast. |
  </Accordion>

  <Accordion title="get_video" icon="clock">
    Returns `{ video: { id, url, status } }`. `status` stays `processing` until the video is ready, then becomes `completed` with a download `url`.

    | Parameter | Type | Description |
    | - | - | - |
    | `video_id` \* | number | Id returned by `generate_video`. |
  </Accordion>
</AccordionGroup>

## Display images and videos

<Accordion title="display_media" icon="images">
  Shows saved images in a gallery or videos in a player in clients that support MCP Apps. Use IDs returned by `generate_image`, `get_image`, `generate_video`, or `get_video`.

  | Parameter | Type | Description |
  | - | - | - |
  | `media_type` \* | string | `image` or `video`. |
  | `ids` \* | integer\[] | One to eight existing IDs owned by your connected account. Duplicate IDs are displayed once. |

  The viewer includes an original-file link. For processing videos, it checks the same ID every 15 seconds for up to five minutes, then offers **Check status**. It never starts a new generation or spends generation credits.

  Clients without embedded UI support still receive the media URLs. Refresh your connector's tool list if `display_media` is missing.
</Accordion>

Image and video tools also return `structuredContent.media`, an array with each result's `id`, `type`, `url`, and `status`. Optional fields include `prompt`, `model`, and `used_credits` when available. Text results remain available for other MCP clients.

## Speech

<AccordionGroup>
  <Accordion title="list_voices" icon="microphone-lines">
    Returns a page of voices with id, name, language, and accent, including your custom and cloned voices. Returns `total`, `offset`, and `next_offset` (`null` on the final page).

    | Parameter | Type | Description |
    | - | - | - |
    | `language` | string | Optional language filter, for example `Dutch`. |
    | `accent` | string | Optional accent filter. |
    | `limit` | integer | Maximum results, default `20`, maximum `50`. |
    | `offset` | integer | Start position, default `0`. Use the previous page's `next_offset`. |
  </Accordion>

  <Accordion title="generate_speech" icon="volume-high">
    Converts text to speech and returns the audio file URL. Billed against the monthly text-to-speech character quota.

    | Parameter | Type | Description |
    | - | - | - |
    | `text` \* | string | Text to speak. |
    | `voice_id` \* | string | Voice id from `list_voices`. |
    | `speed` | number | Playback speed between `0.7` and `1.2` (default `1.0`). |
    | `output_format` | string | For example `mp3_44100_128` (default) or `wav`. |
  </Accordion>
</AccordionGroup>

## Transcription

<AccordionGroup>
  <Accordion title="transcribe_audio" icon="waveform">
    Starts a transcription of an audio or video file and returns a uuid immediately. Requires a paid plan.

    | Parameter | Type | Description |
    | - | - | - |
    | `url` \* | string | Public URL of the audio or video file. |
    | `language` | string | Spoken language name in English, for example `English`. Improves accuracy. |
    | `detect_speakers` | boolean | Label different speakers (default `true`). |
    | `name` | string | Optional name for the transcription. |
  </Accordion>

  <Accordion title="get_transcription" icon="file-lines">
    The `content` field stays empty until the transcription finishes. Reuse the exact UUID and the same connected account. Report a missing or failed job; do not create a replacement unless the user asks for a new transcription.

    | Parameter | Type | Description |
    | - | - | - |
    | `audio_id` \* | string | Transcription uuid from `transcribe_audio`. |
  </Accordion>
</AccordionGroup>

## Async tools

`generate_video` and `transcribe_audio` return an id immediately instead of waiting for the result. Use `display_media` for a video viewer that checks progress, or poll the matching `get_video` or `get_transcription` tool every 15–30 seconds until the result is ready. Always reuse the original ID.

Videos typically take 1–5 minutes depending on model, duration, and resolution.

## Account

<Accordion title="get_account" icon="user">
  Takes no parameters and returns `{ "connected": true }`. Use it only when the user asks to verify their connection. It does not return an account ID, email, name, API key, or OAuth token, and it does not check quota or support account deletion.
</Accordion>


# Create your own tool
Source: https://easy-peasy.ai/docs/tools/custom-tools

Describe a tool in a sentence and get a form with the fields you need, then edit it, reuse it, and share it with your team or anyone.

When no built-in tool fits a job you do often, make your own. Describe it in a sentence; the fields, the instructions, and the form are built for you. You can then adjust it, use it like any other tool, and share it.

## Create a tool

<Steps>
  <Step title="Open Create Tool">
    Click **All Tools** in the sidebar, then **Create Tool** at the top right.
  </Step>

  <Step title="Describe the tool">
    Type what it should make in **What do you want to create?**, such as "yoga class description for a studio schedule". Click **+ add instructions** to say which fields you want and how the result should read.

    <Frame>
      <img alt="Magic AI Tool: Create any AI Tool from a single prompt, with the request yoga class description for a studio schedule, instructions to include fields for the class name, level, length in minutes, and focus, and to write a short, warm description, a Create button, and examples such as ad copy and business plan" />
    </Frame>
  </Step>

  <Step title="Click Create">
    The new tool opens a few seconds later, ready to try.

    <Frame>
      <img alt="The Yoga Class Description Generator tool with Edit and Save buttons, Class Name Slow Flow, Level All Levels, Length 60 minutes, Focus of the Class, Use advanced AI model? off, Examples (optional), outputs 1, language English, and Generate" />
    </Frame>
  </Step>

  <Step title="Try it, then click Save">
    Fill in the fields and click **Generate** to check the results. When you're happy with it, click **Save**. Saved tools appear on the Tools page under **My Tools**.

    <Frame>
      <img alt="A result from the custom tool: Class Name Slow Flow, Level All Levels, Length 60 Minutes, and a class description inviting students to unwind with soothing stretches, breath work, and an extended relaxation, with the usual result buttons" />
    </Frame>

    <Warning>
      A tool you don't save isn't kept. Click **Save** before you leave the page.
    </Warning>
  </Step>
</Steps>

<Frame>
  <img alt="The Tools page filtered to My Tools, showing one tool, Yoga Class Description Generator" />
</Frame>

## Edit a tool

Click **Edit** at the top of the tool.

<Tabs>
  <Tab title="AI Edit">
    Describe the change and click **Apply Changes**. For example: "Remove the Class Description field. Show Level as radio buttons."

    <Frame>
      <img alt="The Edit Template window on AI Edit, with the request Remove the Class Description field. Show Level as radio buttons, and Cancel and Apply Changes" />
    </Frame>
  </Tab>

  <Tab title="Manual Edit">
    Change the **Title**, **Description**, and **Output**, and each field's **Label**, **Type**, **Placeholder**, and **Character Limit**. **Add Field** adds a field, the arrows reorder them, and the trash icon removes one. Click **Save Changes**.

    <Frame>
      <img alt="The Edit Template window on Manual Edit: Title Yoga Class Description Generator, Description, Output Text, Add Field, and Field 1 Class Name as a Text Input with a placeholder and a 100-character limit" />
    </Frame>
  </Tab>
</Tabs>

A tool can have up to 9 fields. Field types are **Text Input**, **Text Area**, **Dropdown Select**, **Checkbox**, **Radio Button**, and **Image Upload**.

**Output** decides what the tool makes:

| Output | What it makes | What it uses |
| - | - | - |
| **Text** | Written text | Words |
| **Image (from description)** | An image from the fields | 3 image & video credits |
| **Image (transform uploaded photo)** | An edited version of an uploaded photo. Needs an **Image Upload** field | 3 image & video credits |
| **Speech (voiced audio)** | A script, read aloud | Words for the script, and text-to-speech characters for the audio |

To delete a saved tool, click the trash icon at the top of it and confirm.

## Use a tool

Your tools work like the built-in ones: fill in the fields, choose the number of results and the language, and click **Generate**. See [Run a tool](/docs/tools/run).

* **Use advanced AI model?** switches text tools to a stronger model, which uses more words.
* **Examples (optional)**: add examples of results you like, and the tool follows their format.
* Results in a language other than English are made one at a time.

## Share a tool

Click **Share** at the top of a saved tool.

<Frame>
  <img alt="The Share Template window for Yoga Class Description Generator with two switches, Share with anyone and Share with your team, both off" />
</Frame>

* **Share with your team** adds the tool to your teammates' **My Tools**. They use it with their own account, and can't edit or delete it.
* **Share with anyone** gives you a link to a public page where anyone can use the tool without an account.

<Warning>
  Results made through a public link use **your** words and credits, and appear in your History. Each person can make up to 20 results a day, and a link up to 300 a day. Visitors always get the standard model and at most 2 results at a time. Turn the switch off to stop the link working.
</Warning>

<Columns>
  <Card title="Run a tool" icon="play" href="/docs/tools/run">
    Options, results, and sharing a result.
  </Card>

  <Card title="Brand Voice" icon="bullhorn" href="/docs/brand-voice/create">
    Make results sound like you.
  </Card>
</Columns>


# Models and words
Source: https://easy-peasy.ai/docs/tools/models

Which AI model to choose in a tool, and how many words from your plan each one uses.

Every tool has an **AI Model** menu. The model you pick is remembered for all tools.

<Frame>
  <img alt="The AI Model menu: Standard model, fast model for everyday tasks, selected; Advanced model, powerful model for complex tasks; Reasoning model, uses advanced reasoning; GPT-6 Astra; GPT-6 Sol; and GPT-6 Luna" />
</Frame>

## Which model to pick

* **Standard model** is the default: fast and good for most everyday writing, such as posts, emails, and descriptions.
* **Advanced model** writes more carefully. Try it for longer or more demanding work, such as reports and proposals.
* **Reasoning model** thinks before it writes. Use it for maths, logic, planning, and step-by-step answers.
* **GPT-6**, **GPT-5.6**, **GPT-5.5**, and **GPT-5 mini** are OpenAI's models, and **Claude Fable**, **Claude Opus**, and **Claude Sonnet** are Anthropic's, for when you prefer a particular model. The Claude options aren't available in a few older tools.

The Claude options marked **Reasoning** think before they write, like the Reasoning model. That thinking counts as writing, so they use more words.

## How many words a result uses

A result uses the words it contains, multiplied by the model's rate.

| Model | Rate on Unlimited and Teams plans |
| - | - |
| **Standard model**, **Reasoning model** | Free |
| **Advanced model**, GPT-5.6 Sol, GPT-5.5, Claude Opus 5.5, Claude Opus 5, Claude Opus 4.8 | 1x |
| GPT-6 Astra, Claude Fable 5.1, Claude Fable 5 | 2x |
| GPT-6 Sol, GPT-5.6 Terra, Claude Sonnet 5, Claude Sonnet 4.6 | 0.5x |
| GPT-5 mini | 0.25x |
| GPT-6 Luna, GPT-5.6 Luna | 0.2x |

On the Free, Basic, and Starter plans, every model counts at 1x, and GPT-6 Astra and Claude Fable at 2x. On the clinical note tools, the **Advanced model** is free on Unlimited and Teams plans, like Standard.

A few things add to the count:

* **Claude models and GPT-6 Astra** also count what they read, such as long inputs and uploaded documents, after the first 1,000 tokens (about 750 words).
* **Reading an uploaded file** adds 150 words per result.
* **More results** use more words: each one counts on its own.

The **History** tab on a tool shows how many words each result used.

## When you run out of words

When your plan's words are used up, tools stop and offer to upgrade. On the Unlimited and Teams plans, the **Standard** and **Reasoning** models keep working. See [Plans and credits](/docs/account/plans-and-credits) and [buy more words](/docs/account/buy-credits).

## On the Free plan

Only the **Standard model** is available. The others show **Upgrade to unlock**.

<Columns>
  <Card title="Run a tool" icon="play" href="/docs/tools/run">
    Fill in the form and work with the results.
  </Card>

  <Card title="Check your usage" icon="chart-simple" href="/docs/account/usage">
    See how many words you've used.
  </Card>
</Columns>


# Tools
Source: https://easy-peasy.ai/docs/tools/overview

Find the right AI writing tool among hundreds, from Instagram captions to lesson plans, and keep your favourites one click away.

Tools are single-purpose AI generators: each one has a short form for one job, such as an Instagram caption, a product description, a cover letter, a lesson plan, or a SOAP note. Fill in the form, click **Generate**, and copy the result.

Use a tool when you do the same kind of writing often and want consistent results. For open-ended work, research, or several steps in a row, use [Marky Agent](/docs/marky/overview).

## Find a tool

Click **All Tools** in the sidebar. There are close to 400, including shortcuts to the other products such as Marky Agent and AI Images.

<Frame>
  <img alt="The Tools page with 394 tools, a Create Tool button, a search box, category filters from Favorites and Writing to Personal & Fun, a Featured sort menu, and the first cards: Marky Agent, AI Art & Images, and AI Note Taker" />
</Frame>

* **Search** by name or by what you want to do, for example "caption" or "email reply".
* **Filter** by category: Writing, Content, Social Media, Marketing, Email & Messages, Business, HR, Career, Education, Images & Video, Audio & Music, Creative Writing, Health & Wellness, Legal, Code & Data, and Personal & Fun.
* **Sort** by **Featured**, **Alphabetical**, or **Newest**. Featured shows your favourites first.

Searching looks through all tools, whatever filter is selected.

You can also find tools from the **Home** screen: **Recommended Tools** shows your favourites first, and its magnifier searches every tool.

## Keep your favourites

Click the star on a tool's card, or at the top right of its form, to make it a favourite. Favourites:

* appear under **★ Favorites** on the Tools page,
* come first in **Featured** and in **Recommended Tools**,
* are added to the sidebar, up to 10.

Click the star again to remove it.

## Your own tools

**Create Tool** builds a new tool from a sentence, with the fields you need. Your tools, and the ones teammates share with you, are under **My Tools**. See [Create your own tool](/docs/tools/custom-tools).

## Tools on the Free plan

On the Free plan, tools use the **Standard model** and make one result at a time. Tools for paid plans can be tried once each: they're marked **Try this template free**, and the button reads **Try Free**. After that one result, the form asks you to upgrade. See [Plans and credits](/docs/account/plans-and-credits#the-free-plan).

<Columns>
  <Card title="Run a tool" icon="play" href="/docs/tools/run">
    Fill in the form and work with the results.
  </Card>

  <Card title="Models and words" icon="microchip" href="/docs/tools/models">
    Pick the AI model and see what it uses.
  </Card>
</Columns>


# Run a tool
Source: https://easy-peasy.ai/docs/tools/run

Fill in a tool's form, choose the model, tone, language, and number of results, and edit, save, or share what it writes.

Every tool works the same way: fill in the form, pick your options, and click **Generate**. This page uses **Instagram Post Caption** as the example.

## Fill in the form

<Frame>
  <img alt="The Instagram Post Caption tool: optional image and video uploads, Post Topic New early-morning flow class, Mondays at 6:30 am, Tone of voice set to the Example Yoga Studio brand voice, Output length Short, AI Model Standard model, Generate Image? on, and Search the web? off" />
</Frame>

<Steps>
  <Step title="Describe what you want">
    Fill in the main field, such as **Post Topic**. Be specific: names, dates, prices, and the one thing readers should remember. Fields with a counter have a length limit.

    The **Dictate** microphone next to a text box types what you say, for up to 10 minutes at a time.
  </Step>

  <Step title="Set the tone">
    **Tone of voice** decides how the result sounds. If you have [brand voices](/docs/brand-voice/create), pick one here; your newest one is selected for you. Pick **Custom** to type a tone instead, such as "witty" or "warm and reassuring". **Manage voices** opens Brand Voice.
  </Step>

  <Step title="Add files (optional)">
    Some tools read an image, a video, or a document you upload, for example a photo to write a caption for, or a PDF to summarize. Reading a file adds 150 words to each result.
  </Step>

  <Step title="Choose the options">
    | Option | What it does |
    | - | - |
    | **Output length** | **Short**, **Medium**, or **Long**, on tools that have it |
    | **AI Model** | Which AI writes the result. See [Models and words](/docs/tools/models) |
    | **Generate Image?** | Adds a matching image to each result, using [image & video credits](/docs/account/plans-and-credits): 3 per image |
    | **Search the web?** | Searches Google first, for results that need current information such as news, prices, or events |
  </Step>

  <Step title="Choose how many results and the language">
    <Frame>
      <img alt="The bar under the form: Clear inputs, outputs 1, language English, and Generate" />
    </Frame>

    **outputs** makes up to 5 different results at once, to compare. **language** writes the result in one of 45 languages, whatever language your inputs are in. **Clear inputs** empties the form.
  </Step>

  <Step title="Click Generate">
    Results appear under **New Outputs** a few seconds later. You can leave the page; results are saved to [History](/docs/get-started/history).
  </Step>
</Steps>

Some tools have two more options:

* **Output Examples (optional)**: paste examples of results you like, and turn them on. The tool follows their format and style.
* **Save default instructions for this tool**: instructions added to every run of that tool, such as your clinic's note format.

## Work with the results

<Frame>
  <img alt="A result: the bottom of a generated image of a sunset yoga class, the caption Sunrise, savasana, and everything in between. Come flow with us Monday mornings at 6:30, your first class is on us, and the buttons thumbs up, thumbs down, delete, Edit, Favorite, Share, Save, and Copy" />
</Frame>

| Button | What it does |
| - | - |
| **Copy** | Copies the result |
| **Save** | Downloads it as a **PDF**, **Word**, or **TXT** file |
| **Edit** | Opens the text in **Edit Content**. Select a part of it, or none, describe the change, such as "make it shorter" or "add emojis", and click **Submit Edit**. The edited version is saved to History as a new result |
| **Favorite** | Adds it to **History** > **Favorites** |
| **Share** | Makes a public link to it. See below |
| Thumbs up and down | Tells us whether the result was useful |
| Trash | Moves it to **History** > **Trash**, where you can restore it |

**History**, next to **New Outputs**, lists everything you've made with this tool.

If a result fails, its card says so and you aren't charged for it. Click **Generate** again.

## Share a result

Click **Share** on a result and turn on **Share with anyone**. You get a link to a public page with the result, and buttons to post it on LinkedIn, Facebook, X, and Reddit. Anyone with the link can view it, and it may be featured as an example on Easy-Peasy.AI.

<Frame>
  <img alt="The Share Output window for an Instagram Post Caption result, with the Share with anyone switch off and the note that anyone with the link can view this output and it may be featured as an example on Easy-Peasy.AI" />
</Frame>

Turn the switch off to make the page private again.

## What it costs

Each result uses words from your plan, based on its length and the model. On the Unlimited and Teams plans, the **Standard model** doesn't use words. **Generate Image?** uses 3 image & video credits per result. See [Models and words](/docs/tools/models).

<Columns>
  <Card title="Models and words" icon="microchip" href="/docs/tools/models">
    Which model to pick, and what it uses.
  </Card>

  <Card title="Create your own tool" icon="wand-magic-sparkles" href="/docs/tools/custom-tools">
    Build a tool for a job you do often.
  </Card>
</Columns>


# Edit a video with AI
Source: https://easy-peasy.ai/docs/videos/edit

Remove, replace, or restyle things in a video by describing the change, for the whole clip or for a few seconds of it.

Describe what should change, such as "remove the person in the background" or "change the lighting to golden evening light", and the rest of the video stays as it is. The edited video keeps the original length, framing, and aspect ratio.

<Note>
  AI video editing is available on paid plans.
</Note>

## Open the editor

* **For a video you made here**: open its **⋮** menu, or its details, and click **Edit with AI**. The **Edit video with AI** window opens.
* **For your own video**: open **Videos** > **Edit Video** and click **Upload a video**: MP4 or MOV, 4 to 30 seconds, up to 100 MB.

## Make an edit

<Frame>
  <img alt="The Edit video with AI window: the video with a play button and Add edit here, a timeline with edit 2 covering 0 to 2 seconds and edit 1 covering the whole clip; on the right, edit 1 Whole clip 'Change the lighting to golden evening light with long warm shadows' and edit 2 0s – 2s 'Add a rolled-out purple yoga mat in the middle of the floor', Quick edits, and the Seedance 2.5 model; at the bottom 720p, 75 credits, Cancel, and Apply edits" />
</Frame>

<Steps>
  <Step title="Describe the change">
    Type the change in the first edit, which applies to the **Whole clip**. Start with what you want done: add, remove, replace, or change. **Quick edits** start a sentence for you: **Remove an object**, **Replace a subject**, **Change the background**, **Change the lighting**, **Change the outfit**, **Remove on-screen text**, **Remove the music**, and **Change the style**.
  </Step>

  <Step title="Limit an edit to a moment (optional)">
    Play the video, pause where the change should happen, and click **Add edit here**. The new edit covers a few seconds from that point. Drag its ends on the timeline, or use **start here** and **end here** while the video plays. Times snap to whole seconds, so treat them as a strong hint rather than an exact cut. You can add up to 6 edits.
  </Step>

  <Step title="Add reference images (optional)">
    Under **References (optional)**, add up to 4 images, such as a jacket or a person to swap in. They're tagged **@Image1** to **@Image4**, so you can write "replace the jacket with @Image1".
  </Step>

  <Step title="Pick an engine and resolution">
    Choose a model under **Model**, and a resolution at the bottom. The price updates as you go.
  </Step>

  <Step title="Click Apply edits">
    The edited video appears in your results in a few minutes. Its menu says **Change these edits** instead of **Regenerate**, so you can adjust the instructions and try again.
  </Step>
</Steps>

## Engines and credits

Editing is charged per second of the video you edit, at the resolution you choose. Seconds are rounded up.

| Model | Good for | Video length | Credits per second |
| - | - | - | - |
| **Seedance 2.5** (default) | Best quality | 4–30 s | 7 at 480p, 15 at 720p |
| **Seedance 2.5 Fast** | Faster and cheaper | 4–30 s | 9 at 720p, 10 at 1080p |
| **WAN 3.0** | Another model family: try it when Seedance refuses a clip | 4–15 s | 5 at 480p, 10 at 720p, 20 at 1080p |

For example, editing a 5-second video with Seedance 2.5 at 720p costs 75 credits.

## Tips

* Keep each edit to one change. Use separate edits for separate changes.
* Name what's there and what should replace it: "replace the red car with a white van" works better than "change the car".
* If an edit is refused, which can happen with clips that show real people, try **WAN 3.0**. You aren't charged for refused edits.

<Columns>
  <Card title="Extend a video" icon="forward" href="/docs/videos/extend">
    Add a new shot to the end.
  </Card>

  <Card title="Upscale a video" icon="up-right-and-down-left-from-center" href="/docs/videos/upscale">
    Make the result sharper.
  </Card>
</Columns>


# Extend a video
Source: https://easy-peasy.ai/docs/videos/extend

Continue a video from its last frame, trim it first if needed, and get one longer video with the original and the new part.

Extending adds a new shot to the end of a video. The new part starts from the last frame, and you get back one video: the original plus the continuation. Extend the result again to keep going.

<Note>
  Extending videos is available on paid plans.
</Note>

## Extend a video

<Steps>
  <Step title="Choose the video">
    On one of your videos, open the **⋮** menu and click **Extend video**. Or open **Videos** > **Extend Video** and click **Upload a video** to use your own: MP4 or MOV, 2 to 30 seconds, up to 100 MB.

    <Frame>
      <img alt="Extend Video with a 4-second yoga studio clip under Video to extend, a Trim this video link, the note that the new segment starts from the last frame, a What happens next? prompt, the first model card, and an Extend (12 credits) button" />
    </Frame>
  </Step>

  <Step title="Trim the end (optional)">
    The continuation starts from the very last frame, so cut off any part you don't want first. Click **Trim this video**, drag the handles to choose what to keep, and click **Apply trim**. Dragging the end handle shows the exact frame the new part will start from. Trimming is free, and **Undo trim** brings the full video back.

    <Frame>
      <img alt="The trim controls under the video: Keep 0:00.0 – 0:03.6, a 3.6s length, a range bar with two handles, and Apply trim and Cancel buttons" />
    </Frame>
  </Step>

  <Step title="Describe what happens next">
    In **What happens next?**, describe the new part only: the action, the camera move, and the mood. For example: "The camera keeps gliding forward to the window as the sun rises higher and the room fills with warm light."
  </Step>

  <Step title="Pick a model, length, and resolution">
    Choose a model, how many seconds to **Add**, and a **Resolution**. **Finished video** shows the total length.

    <Frame>
      <img alt="Model cards Seedance 2.0 Fast Extend (12 credits), Seedance 2.0 Mini Extend (8), Seedance 2.0 Extend (24), Seedance 2.5 Extend (50, Bills the original clip too), and Gemini Omni Flash 1.1 Extend (18, Bills the original clip too); Add buttons from 4s to 15s; Finished video ~8s; Resolution 480p, 720p, 1080p, 4K" />
    </Frame>
  </Step>

  <Step title="Click Extend">
    The button shows the price, for example **Extend (12 credits)**. The longer video appears in your results in a few minutes.
  </Step>
</Steps>

## Models and credits

| Model | Adds | Longest video you can extend | Credits per second at 720p |
| - | - | - | - |
| **Seedance 2.0 Fast Extend** (default) | 4–15 s | 15 s | 6 for each second added |
| **Seedance 2.0 Mini Extend** | 4–15 s | 15 s | 4 for each second added |
| **Seedance 2.0 Extend** | 4–15 s | 15 s | 12 for each second added |
| **Seedance 2.5 Extend** | 4–30 s | 30 s | 11 for each second of the original and the new part |
| **Gemini Omni Flash 1.1 Extend** | 3–10 s | 10 s, up to 20 MB | 6 for each second of the original and the new part |

* **480p costs half, and higher resolutions cost more**: 1080p is 2.5 times the 720p price and 4K is 5 times on Seedance. On Gemini Omni Flash, 1080p is 1.5 times and 4K is 3 times, and a 360p tier costs a third.
* **The Seedance 2.0 models charge only for the seconds you add.** Seedance 2.5 and Gemini Omni Flash also charge for the original video, marked **Bills the original clip too**. Their note under the resolution shows the split, such as "Billed for 9s — 5s original + 4s new". Trimming the original first makes these cheaper.
* Gemini Omni Flash continues from up to 10 seconds of the video rather than the last frame alone, and makes the whole video again at the resolution you choose.

## If something goes wrong

* **"Could not read this video's length"**: only MP4 and MOV videos can be extended. Export the video again as MP4 and upload it.
* **The video is too long for the model**: trim it, or pick a model that takes longer videos, such as Seedance 2.5 Extend.
* If the extension fails, you aren't charged.

<Columns>
  <Card title="Edit a video with AI" icon="pen" href="/docs/videos/edit">
    Change something in a video instead of adding to it.
  </Card>

  <Card title="Upscale a video" icon="up-right-and-down-left-from-center" href="/docs/videos/upscale">
    Make the result sharper.
  </Card>
</Columns>


# Generate a video
Source: https://easy-peasy.ai/docs/videos/generate

Describe a scene or animate an image, pick a model, set the length, shape, and quality, and generate a video with sound.

Describe what should happen, and a video model turns it into a clip, usually with sound. You can also start from one of your images and describe how it should move.

<Frame>
  <img alt="The left panel of the Video Generator: References with Add media, Start image and End image, a prompt about a sunlit yoga studio, the model list with MiniMax H3 selected, and a bottom bar with 4s, 16:9, 768p, and Generate (12 credits)" />
</Frame>

<Steps>
  <Step title="Describe the video">
    Type what you want to see in **Describe your video**: the subject, what happens, the setting, the light, and how the camera moves. For example: "Slow dolly shot through a sunlit yoga studio at sunrise: warm wooden floor, rolled mats, tall windows, potted plants, dust drifting in the beams of light. Calm, cinematic, no people."

    The counter under the box shows how long your prompt can be for the selected model: 4,000 characters for MiniMax H3, and up to 20,000 for Seedance 2.5 and WAN 3.0.

    Need help with the wording? **Generate prompt** has two options. **Simple** writes a prompt for you, or a random one when the box is empty. **Advanced** asks about the concept, style, camera, pacing, and effects first. Both use a few words from your plan.
  </Step>

  <Step title="Start from an image (optional)">
    Click **Start image** to animate one of your images. The label above the prompt changes to **How do you want to animate this image?**, so describe the motion rather than the picture. Add an **End image** too, and the video moves from the first image to the last.

    Images must be at least 512 × 512 pixels. You can pick an upload or any image from your **History**, so you can make the perfect first frame in [AI Images](/docs/ai-images/generate) and animate it here.

    To keep a person or product the same without fixing the first frame, use [references](/docs/videos/references) instead. A video can have a start image or references, not both.
  </Step>

  <Step title="Pick a model">
    The list under **Model** shows the models that fit what you've added: text models when there's no image, image models when there's a start image, and models that support a start and end image when you add both. If your model can't take what you added, the page switches to one that can, so check the model before you generate.

    **MiniMax H3** is selected on paid plans. Prices aren't shown in the list, but the **Generate** button updates as you switch. See [Video models and credits](/docs/videos/models) to compare them.
  </Step>

  <Step title="Set the length, shape, quality, and sound">
    The bar at the bottom holds the settings for the selected model:

    <Frame>
      <img alt="The Duration menu open above the bottom bar, listing 4 seconds (selected), 5, 6, 7, 8, and 9 seconds" />
    </Frame>

    | Setting | What it does |
    | - | - |
    | Duration | The length in seconds. Each model offers its own lengths, from 3 seconds up to 30 with Seedance 2.5 and WAN 3.0. |
    | Aspect ratio | The shape, such as **16:9** for YouTube or **9:16** for Reels and TikTok. Shown only when the model offers more than one. Image models usually follow the shape of your start image. |
    | Resolution | The quality, such as **480p**, **720p**, **1080p**, **2K**, or **4K**. Higher resolutions cost more. |
    | **Audio** | Turns the soundtrack on or off. On Veo 3.1, Veo 3.1 Fast, Kling 3.0, Kling O3, and Seedance 1.5 Pro, a silent video costs less, and the switch shows the discount. MiniMax H3 and Gemini Omni Flash 1.1 always add sound, so they don't show it. |
  </Step>

  <Step title="Click Generate">
    The button shows the total, for example **Generate (12 credits)**. Your video appears in the **Videos** tab as **Processing video...** and is usually ready in 2 to 5 minutes. The list updates by itself. You can leave the page: the video keeps rendering and appears in **History** > **Videos** when it's done.

    <Frame>
      <img alt="A placeholder tile in the Videos tab with a spinner, Processing video..., and This can take 2-5 minutes" />
    </Frame>
  </Step>
</Steps>

**Clear inputs** empties the prompt, images, references, and scene settings. It keeps the model, length, and aspect ratio.

## Add camera, lighting, and shot directions

Click **Add media** and use the **Scene** tabs, **Camera**, **Lighting**, and **Shot Type**, to add ready-made directions such as a lens, **Golden Hour** light, or a **Low Angle** shot. They're added to the end of your prompt when you generate. They work the same way as in [AI Images](/docs/ai-images/generate).

## Tips for better videos

* Describe one continuous shot: who or what is in it, what happens, how the camera moves, and the light.
* Say what should be heard. Models that make sound follow directions such as "soft piano music", "birds outside", or dialogue in quotes.
* For full control of the look, make the first frame in AI Images, then use it as the **Start image**.
* Test cheaply first: 4 seconds at 480p on Seedance 2.0 Mini costs 8 credits. Switch to a stronger model or a higher resolution once the idea works.
* To keep the same character across several videos, add them as a [reference](/docs/videos/references) every time.

## On the Free plan

* You can use **Seedance 2.0 Mini** (480p, up to 5 seconds), and **Veo 3.1 Fast** and **Veo 3.1 Lite** (720p). Other models show a crown and can't be selected.
* Videos use the same credits as images, and the shortest video costs 8 credits, so you may need [bonus credits](/docs/account/buy-credits) or an upgrade to make one.
* Extending, editing, and upscaling videos, and Motion Control, need a paid plan.

## If a video fails

* A tile that says **Generation Failed** doesn't cost anything. The video was refused, usually because of the prompt or an image, or the model couldn't finish it. Try again with a different prompt or another model.
* Sometimes a failed video disappears from the list instead. Its credits are returned too.
* Images you add are checked when you add them. If one breaks the content policy, it's marked **Blocked** and the button changes to **Content Blocked**. Remove it to continue.

<Columns>
  <Card title="Use references" icon="user" href="/docs/videos/references">
    Keep a character or product the same across videos.
  </Card>

  <Card title="Manage your videos" icon="folder-open" href="/docs/videos/manage">
    Download, share, extend, or edit your results.
  </Card>
</Columns>


# Manage your videos
Source: https://easy-peasy.ai/docs/videos/manage

Find, download, share, rate, and delete your videos, and get credits back for a result you're not happy with.

## Where your videos are

* **The Videos tab** on the Video Generator shows your latest videos. On a team, it also shows videos your teammates shared with the team.
* **History** > **Videos** has every video you've made: generated videos, talking videos, edits, extensions, upscales, and Video Editor exports. Search by prompt or title, filter by date, and on a team, filter by person.

<Frame>
  <img alt="History with the Videos tab selected, a search box for prompts or titles, a date range filter, and three videos: a Video Editor export, a talking video, and a generated video of a yoga studio" />
</Frame>

Hover over a video to play it without sound. Click it to open its details.

## The video menu

Each video has a **⋮** menu:

<Frame>
  <img alt="The menu of a video: Open details, Edit with AI, Upscale video, Upgrade to 2K · MiniMax, Extend video, Regenerate, Download, Share, Good video, Bad video, and Delete" />
</Frame>

| Item | What it does |
| - | - |
| **Open details** | Shows how the video was made. |
| **Edit with AI** | [Edit the video](/docs/videos/edit) by describing a change. |
| **Upscale video** | [Upscale the video](/docs/videos/upscale), up to 4K. |
| **Upgrade to 2K · MiniMax** | Remake a 768p MiniMax H3 video at 2K. See [Upscale a video](/docs/videos/upscale#upgrade-a-minimax-h3-video-to-2k). |
| **Extend video** | [Continue the video](/docs/videos/extend) from its last frame. |
| **Regenerate** | Reopen the tool with the same prompt, model, and settings, so you can make another version. For edits, extensions, and upscales it says **Change these edits**, **Extend again**, or **Upscale again**. |
| **Download** | Save the video as an MP4. |
| **Share** | Get a public link, or share with your team. |
| **Good video**, **Bad video** | Rate the result. **Bad video** also returns its credits. |
| **Delete** | Delete the video. |

## Video details

**Open details** shows the model, the credits the video used, the prompt (with a button to copy it), and the date. The same actions are here as buttons.

<Frame>
  <img alt="The details of a video: model MiniMax H3, 12 credits used, the prompt about a sunlit yoga studio, the creation date, a Share with Team switch, and buttons for Edit with AI, Upscale video, Upgrade to 2K · MiniMax, Extend video, Regenerate, Download video, and Video without sound effects, next to the video" />
</Frame>

Some videos also have **Video without sound effects**, which opens the version before sound was added.

## Share a video

* **Share with Team** in the details shows the video to everyone on your team.
* **Share** opens **Share Video**. Turn on **Share Publicly** to get a public page for the video on easy-peasy.ai, with buttons to post it to LinkedIn, Facebook, X, and Reddit. Turn it off to take the page down. **Share with Team Members** is available on Teams plans.

Videos stay private unless you share them.

## Get credits back for a bad result

If a video came out wrong, open its menu and click **Bad video**. You get its credits back, up to 10% of your plan's monthly credits in each billing period. It works for your own videos from the current billing period. Your feedback also helps us improve the models.

## Delete a video

Open the menu, click **Delete**, and confirm. Deleted videos can't be restored, and deleting doesn't return the credits they used.

## Download

**Download** saves the video as an MP4, with no watermark.

<Columns>
  <Card title="Generate a video" icon="clapperboard" href="/docs/videos/generate">
    Make a new video.
  </Card>

  <Card title="Plans and credits" icon="coins" href="/docs/account/plans-and-credits">
    How image & video credits work.
  </Card>
</Columns>


# Video models and credits
Source: https://easy-peasy.ai/docs/videos/models

Every video model on Easy-Peasy.AI, what it can start from, the lengths and resolutions it offers, what it costs per second, and which ones the Free plan can use.

You choose a model under **Model** on the [Video Generator](/docs/videos/generate). This page lists the same models, grouped by family, in the same order as the list.

## How prices work

* **Per second.** Most models charge per second of video, so a 10-second video costs twice as much as a 5-second one. A few older models charge a set price per video.
* **By resolution.** Higher resolutions cost more. The table shows the price per second at each resolution.
* **Sound.** Most models make sound with the video. On some, a silent video costs less, as shown in the table.
* **References.** Reference videos add to the price on some models. See [Use references](/docs/videos/references#what-references-cost).
* **Shown before you start.** The **Generate** button shows the total for your settings. Totals are rounded up to whole credits.

The default model on paid plans is **MiniMax H3**: a 4-second video at 768p costs 12 credits.

## Video models

"Starts from" lists what each family can make a video from. A family's image and reference models appear in the list when you add a start image or references; for example, **Seedance 2.5** becomes **Seedance 2.5 Image** or **Seedance 2.5 Reference**.

| Model | Starts from | Length | Quality | Credits | Free plan |
| - | - | - | - | - | - |
| MiniMax H3 | Text, image, start and end images, reference images and video | 4–15 s | 768p, 2K | 3 per second at 768p, 5 at 2K | — |
| Seedance 2.0 Fast | Text, image, start and end images, reference images and video | 4–15 s | 480p, 720p, 1080p, 4K | 3 per second at 480p, 6 at 720p, 15 at 1080p, 30 at 4K | — |
| Seedance 2.0 Mini | Text, image, start and end images, reference images and video | 4–15 s | 480p, 720p, 1080p, 4K | 2 per second at 480p, 4 at 720p, 10 at 1080p, 20 at 4K | 480p, up to 5 s |
| Seedance 2.5 | Text, image, start and end images, reference images and video | 4, 5, 6, 8, 10, 12, 15, 20, 25, 30 s | 480p, 720p | 6 per second at 480p, 12 at 720p | — |
| Seedance 2.5 Turbo | Text, image, start and end images, reference images and video | 4, 5, 6, 8, 10, 12, 15, 20, 25, 30 s | 720p, 1080p | 7 per second at 720p, 7.35 at 1080p | — |
| FLUX 3 | Text, image, start and end images | 5, 6, 8, 10, 12, 15, 20 s | 720p, 1080p | 9 per second at 720p, 15.3 at 1080p | — |
| WAN 3.0 | Text, image, start and end images, reference images | 5, 8, 10, 12, 15, 20, 25, 30 s | 480p, 720p, 1080p | 2.5 per second at 480p, 5 at 720p, 10 at 1080p | — |
| Veo 3.1 Fast | Text, image, start and end images, reference images | 4, 6, 8 s (8 s from an image) | 720p, 1080p, 4K | 5 per second at 720p, 6 at 1080p, 15 at 4K (15% less without sound) | 720p |
| Gemini Omni Flash | Text, image, reference images and video | 4–10 s | — | 6 per second | — |
| Gemini Omni Flash 1.1 | Text, image, start and end images, reference images and video | 3–10 s | 360p, 720p, 1080p, 4K | 2 per second at 360p, 6 at 720p, 9 at 1080p, 18 at 4K | — |
| MiniMax H3 Max | Text, image, start and end images, reference images and video | 5–15 s | 480p, 768p | 2 per second at 480p, 4 at 768p; MiniMax H3 Max Reference: 4 per second at 480p, 4 at 768p | — |
| Grok Imagine | Text, image, reference images | 5–15 s | 480p, 720p | 2 per second at 480p, 3 at 720p | — |
| Grok Imagine 1.5 | Text, image, reference images | 5–15 s | 480p, 720p, 1080p | 4 per second at 480p, 7 at 720p, 12.5 at 1080p | — |
| Seedance 2.0 Turbo | Text, image, start and end images | 4–15 s | 720p, 1080p | 7 per second at 720p, 7.49 at 1080p | — |
| Seedance 2.0 | Text, image, start and end images, reference images and video | 4–15 s | 480p, 720p, 1080p, 4K | 6 per second at 480p, 12 at 720p, 30 at 1080p, 60 at 4K | — |
| Kling 2.6 Pro | Text, image, start and end images | 5, 10 s | — | 30 for 5 s, 60 for 10 s | — |
| Veo 3.1 Lite | Text, image | 4, 6, 8 s (8 s from an image) | 720p, 1080p | 2.5 per second at 720p, 4 at 1080p | 720p |
| Happy Horse | Text, image, reference images | 3–15 s | 720p, 1080p | 7 per second at 720p, 14 at 1080p | — |
| Kling 3.0 Pro | Text, image, start and end images | 3–15 s | — | 15 per second (25% less without sound) | — |
| Kling 3.0 Standard | Text, image, start and end images | 3–15 s | — | 10 per second (25% less without sound) | — |
| Kling O3 Pro | Text, image, start and end images | 3–15 s | — | 7 per second (25% less without sound) | — |
| Kling O3 Standard | Text, image, start and end images | 3–15 s | — | 5 per second (25% less without sound) | — |
| Kling O3 Reference | Reference images | 3–15 s | — | 5 per second (25% less without sound) | — |
| Seedance 1.5 Pro | Text, image, start and end images | 4–12 s | 480p, 720p, 1080p | 2 per second at 480p, 4 at 720p, 8 at 1080p (50% less without sound) | — |
| Kling O1 | Image, start and end images, reference images | 3, 5, 10 s (5, 10 s from an image) | — | 25 for 5 s, 50 for 10 s | — |
| Kling 2.5 Turbo Pro | Text, image, start and end images | 5, 10 s | — | 18 for 5 s, 36 for 10 s | — |
| Kling 2.5 Turbo Standard | Image | 5, 10 s | — | 10 for 5 s, 20 for 10 s | — |
| Veo 3.1 | Text, image, start and end images, reference images | 4, 6, 8 s (8 s from an image) | 720p, 1080p, 4K | 20 per second at 720p, 20 at 1080p, 30 at 4K (50% less without sound) | — |
| LTX-2 Pro | Text, image | 6, 8, 10 s | — | 3 per second | — |
| LTX-2 Fast | Text, image | 6, 8, 10, 12, 14, 16, 18, 20 s | — | 2 per second | — |
| Hailuo 2.0 | Text, image, start and end images | 6, 10 s | — | 15 for 6 s, 30 for 10 s | — |
| Hailuo 2.0 Pro | Text, image, start and end images | Set by the model | — | 25 per video | — |
| Hailuo 2.3 | Text, image | 6, 10 s | — | 14 for 6 s, 28 for 10 s | — |
| Hailuo 2.3 Pro | Text, image | 5 s | — | 24 for 5 s | — |
| Hailuo 2.3-Fast Pro | Image | 6 s | — | 16 for 6 s | — |
| Seedance v1 Pro Fast | Text, image | 3–12 s | — | 1.5 per second | — |
| Pixverse v5.5 | Text, image | 5, 10 s | — | 15 for 5 s, 30 for 10 s | — |
| Wan 2.5 | Text, image | Set by the model | — | 35 per video | — |
| Wan v2.2 Turbo | Text, image, start and end images | Set by the model | — | 5 per video | — |
| Wan v2.2 | Text, image, start and end images | Set by the model | — | 20 per video | — |

## Picking a model

* **To try an idea cheaply**: Seedance 2.0 Mini at 480p, 2 credits per second.
* **For long takes**: Seedance 2.5 and WAN 3.0 make up to 30 seconds in one video.
* **For a person or product that stays the same**: a model that takes reference images, such as Seedance 2.5 Reference.
* **For speed**: MiniMax H3 Max renders in seconds rather than minutes.

Other video tools have their own prices: see [Extend a video](/docs/videos/extend), [Edit a video with AI](/docs/videos/edit), [Upscale a video](/docs/videos/upscale), [Motion Control](/docs/videos/motion-control), and [Talking videos](/docs/videos/talking-videos).


# Motion Control
Source: https://easy-peasy.ai/docs/videos/motion-control

Copy the movement, gestures, and timing from a reference video onto a character in a photo.

Motion Control takes the movement from one video, such as a dance, a gesture, or a walk, and applies it to the character in an image. The character moves exactly like the person in your reference video.

<Note>
  Motion Control is available on paid plans.
</Note>

## Make a video

<Steps>
  <Step title="Open Motion Control">
    Choose **Videos** > **Motion Control**. **How it works** at the top shows an example.
  </Step>

  <Step title="Add the reference video and the character">
    * **Upload reference video**: the movement to copy. MP4 or MOV, 3 to 30 seconds, up to 100 MB.
    * **Add character image**: the character to animate. JPEG or PNG, at least 512 × 512 pixels, up to 10 MB.

    For the best results, match the framing: a full-body video with a full-body image, or half-body with half-body.

    <Frame>
      <img alt="Motion Control with Upload reference video (Duration required: 3-30 secs) and Add character image tiles, an optional prompt, Orientation Video or Image, Quality 1080p Pro or 720p, a Voice Change checkbox, and a Generate (15 credits) button" />
    </Frame>
  </Step>

  <Step title="Describe the scene (optional)">
    The movement comes from the video, so use **Prompt** for everything else: the background, the setting, extra details. For example: "A corgi runs in" or "Snowy park setting".
  </Step>

  <Step title="Choose the settings">
    | Setting | Options |
    | - | - |
    | **Orientation** | Which way the character faces. **Video** follows the reference video. **Image** keeps the direction from your image, and lets you add camera movement in the prompt. |
    | **Quality** | **1080p** (**Pro**, the default) or **720p**. |
    | **Voice Change** | Turn on **Enable voice change** and pick a voice to change the voice in the finished video. |
  </Step>

  <Step title="Click Generate">
    The button shows the price. The video appears in your results in a few minutes.
  </Step>
</Steps>

## Tips

* Use steady, moderate movements. Avoid fast action and camera cuts in the reference video.
* Keep the character fully visible in both the image and the video.

## Credits

Motion Control is charged per second of the reference video, rounded up: **5 credits per second at 1080p** and **2.5 at 720p**. A 10-second reference video costs 50 credits at 1080p. Until you add a video, the button shows the price for 3 seconds.

<Columns>
  <Card title="Use references" icon="user" href="/docs/videos/references">
    Guide the camera and motion of a new video with a clip.
  </Card>

  <Card title="Talking videos" icon="comment" href="/docs/videos/talking-videos">
    Make a character speak instead.
  </Card>
</Columns>


# AI Videos
Source: https://easy-peasy.ai/docs/videos/overview

Generate videos from text or images, extend and edit them, make talking videos, and build a full edited video from one sentence.

<Frame>
  <img alt="The Video Generator: references, start and end images, a prompt about a sunlit yoga studio, the model list with MiniMax H3 selected, and a Generate (12 credits) button on the left; three finished videos on the right" />
</Frame>

Open **AI Videos** in the sidebar, or choose a tool from the **Videos** menu at the top of any tool page. The Video Generator has two halves:

* **On the left**, you describe the video, add a start image or references if you want them, pick a model, and click **Generate**.
* **On the right**, you see your results and ideas to start from.

| Tab on the right | What's in it |
| - | - |
| **Videos** | Your latest videos. Everything older is in **History** > **Videos**. |
| **Templates** | Ready-made videos you can remake, many of them with you in the starring role. |
| **Examples** | Videos other people chose to share publicly, for inspiration. |

## What you can do

<Columns>
  <Card title="Generate a video" icon="clapperboard" href="/docs/videos/generate">
    Describe a scene or animate an image, then pick a model, length, and quality.
  </Card>

  <Card title="Use references" icon="user" href="/docs/videos/references">
    Keep a person, product, or look the same, and guide the motion or the sound.
  </Card>

  <Card title="Video templates" icon="shapes" href="/docs/videos/templates">
    Start from a ready-made video and put your own photo in it.
  </Card>

  <Card title="Extend a video" icon="forward" href="/docs/videos/extend">
    Continue a clip from its last frame, and trim it first if you need to.
  </Card>

  <Card title="Edit a video with AI" icon="pen" href="/docs/videos/edit">
    Remove, replace, or restyle things in a video by describing the change.
  </Card>

  <Card title="Upscale a video" icon="up-right-and-down-left-from-center" href="/docs/videos/upscale">
    Make a video sharper, up to 4K.
  </Card>

  <Card title="Motion Control" icon="person-running" href="/docs/videos/motion-control">
    Copy the movement from one video onto a character in a photo.
  </Card>

  <Card title="Talking videos" icon="comment" href="/docs/videos/talking-videos">
    Make a photo, an AI actor, or a video speak your script.
  </Card>

  <Card title="Video Agent" icon="robot" href="/docs/videos/video-agent">
    Turn one sentence into a finished video with voiceover, music, and captions.
  </Card>

  <Card title="Video Editor" icon="film" href="/docs/videos/video-editor">
    Arrange clips, captions, and music on a timeline, then export an MP4.
  </Card>

  <Card title="Manage your videos" icon="folder-open" href="/docs/videos/manage">
    Download, share, rate, and find everything you've made.
  </Card>

  <Card title="Video models and credits" icon="table" href="/docs/videos/models">
    Every video model, what it can start from, and what it costs.
  </Card>
</Columns>

## The Videos menu

The **Videos** menu at the top of the page opens every video tool:

| Menu item | What it opens |
| - | - |
| **Video Agent** | The [Video Agent](/docs/videos/video-agent), which plans, voices, and assembles a whole video. |
| **Video Generator** | The main page, described in [Generate a video](/docs/videos/generate). |
| **Motion Control** | [Motion Control](/docs/videos/motion-control). |
| **Extend Video**, **Edit Video**, **Upscale Video** | The [extend](/docs/videos/extend), [edit](/docs/videos/edit), and [upscale](/docs/videos/upscale) tools. |
| **Talking Videos** | [Talking videos](/docs/videos/talking-videos). |
| **Video Editor** | The [Video Editor](/docs/videos/video-editor). |
| **Seedance 2.5**, **MiniMax H3**, **WAN 3.0**, **Gemini Omni Flash 1.1** | The Video Generator with that model selected. |

## Credits

Videos use image & video credits, the same pool as images. Most models charge per second of video, and the price also depends on the resolution: a 4-second MiniMax H3 video at 768p costs 12 credits. The **Generate** button shows the total before you start. See [Video models and credits](/docs/videos/models) for every model's price, and [Plans and credits](/docs/account/plans-and-credits) for how credits work.

On the Free plan, only a few models are available at short lengths and low resolutions, and extending, editing, upscaling, and Motion Control need a paid plan.


# Use references
Source: https://easy-peasy.ai/docs/videos/references

Add images of a person, product, or style, a video to guide the motion, or audio to guide the sound, and keep them consistent across videos.

References tell the model what should look or sound the same, without fixing the first frame the way a start image does. Use them to put the same person or product in every video, match a look, copy the camera movement from a clip, or give a character a particular voice.

| Reference | What it's for |
| - | - |
| **Images** | People, products, places, or a style to match. A **Character** or **Object** from your library keeps the same person or product across videos. |
| **Video** | A clip whose motion, camera work, and pacing the new video follows. On Gemini Omni Flash, the clip is the video you want to edit instead. |
| **Audio** | A voice, song, or piece of music that guides the soundtrack. It always goes with at least one image or video reference. |

## Add references

<Steps>
  <Step title="Open Add media">
    Click **Add media** under **References (optional)**.

    <Frame>
      <img alt="The Add media window: Library tabs All uploads, Character, Object, History, Video, and Audio; Scene tabs Camera, Lighting, and Shot Type; an Upload new tile for images, video and audio; and Clear selection, Cancel, and Apply buttons" />
    </Frame>
  </Step>

  <Step title="Pick your media">
    Choose from the **Library** tabs:

    * **All uploads**: everything you've uploaded. **Upload new** takes images, MP4 or MOV videos, and audio files.
    * **Character** and **Object**: the people and products you've saved in your library, also used in [AI Images](/docs/ai-images/generate).
    * **History**: images you've generated.
    * **Video** and **Audio**: a reference clip and a sound file. If an audio file is longer than the model allows, drag the highlighted window over its waveform to choose the part to use.

    You can attach one video and one audio file per video.
  </Step>

  <Step title="Click Apply">
    The references appear under **References (optional)** with a counter, such as **1 / 30**. If your model can't use references, the page switches to its reference version, for example from **Seedance 2.0 Fast** to **Seedance 2.0 Fast Reference**. Remove the references and it switches back.
  </Step>

  <Step title="Say how to use them in your prompt">
    Describe what each reference is for. Many models understand **@image1**, **@image2**, and so on, in the order you added them: "@image1 is the main character. She walks into a sunlit yoga studio and rolls out her mat."
  </Step>
</Steps>

<Frame>
  <img alt="A reference labelled Character under References (optional) with a 1 / 30 counter, and a prompt that begins Create a 15-second cinematic surreal photorealistic video. @image1 is the main character" />
</Frame>

## Rules

* **References or a start image, not both.** Adding a start or end image removes the references, and adding references hides the start and end images.
* **Audio needs company.** Add at least one image or video reference with any audio reference.
* **Audio formats.** MiniMax H3, Seedance 2.5, and WAN 3.0 accept only WAV or MP3 audio.
* **Video files** must be MP4 or MOV.
* Images you add are checked against the content policy. A blocked image is marked **Blocked** and must be removed before you can generate.

## Limits by model

The number of references each model accepts per video. Models that aren't listed don't take references.

| Model | Reference images | Reference video | Reference audio |
| - | - | - | - |
| MiniMax H3 | Up to 9 | Up to 3 clips, 2–15 s each, 15 s in total | Up to 3 clips, 15 s in total, WAV or MP3 |
| Seedance 2.0 Fast | Up to 9 | Up to 3 clips, 2–15 s each, 15 s in total | Up to 3 clips, 15 s in total |
| Seedance 2.0 Mini | Up to 9 | Up to 3 clips, 2–15 s each, 15 s in total | Up to 3 clips, 15 s in total |
| Seedance 2.5 | Up to 30 | Up to 3 clips, 2–30 s each, 30 s in total | Up to 10 clips, 30 s in total, WAV or MP3 |
| Seedance 2.5 Turbo | Up to 30 | Up to 3 clips, 2–30 s each, 30 s in total | Up to 10 clips, 30 s in total, WAV or MP3 |
| WAN 3.0 | Up to 9 | — | Up to 3 clips, 30 s in total, WAV or MP3 |
| Veo 3.1 Fast | Up to 3 (people and objects) | — | — |
| Gemini Omni Flash | Up to 5 | 1 clip to edit, 4–20 s | — |
| Gemini Omni Flash 1.1 | Up to 5 | 1 clip to edit, 4–10 s | — |
| MiniMax H3 Max | Up to 9 | Up to 3 clips, 2–15 s each, 15 s in total | Up to 3 clips, 15 s in total |
| Grok Imagine | Up to 7 | — | — |
| Grok Imagine 1.5 | Up to 7 | — | — |
| Seedance 2.0 | Up to 9 | Up to 3 clips, 2–15 s each, 15 s in total | Up to 3 clips, 15 s in total |
| Happy Horse | Up to 9 | — | — |
| Kling O3 Reference | Up to 4 | — | — |
| Kling O1 | Up to 4 | — | — |
| Veo 3.1 | Up to 3 (people and objects) | — | — |

## What references cost

* **Images** are included in the price on most models.
* **Reference videos** are charged by their length on Seedance 2.0, Seedance 2.5, and MiniMax H3, at the same rate per second as the video you generate. For example, a 10-second clip adds 60 credits to a Seedance 2.5 Reference video at 480p.
* On **MiniMax H3 Max Reference**, images and clips are charged by size: about four images are free, and each clip adds credits by its length.

The **Generate** button includes all of this before you start.

<Columns>
  <Card title="Video templates" icon="shapes" href="/docs/videos/templates">
    Many templates are built around one reference photo of you.
  </Card>

  <Card title="Video models and credits" icon="table" href="/docs/videos/models">
    Compare the models that take references.
  </Card>
</Columns>


# Talking videos
Source: https://easy-peasy.ai/docs/videos/talking-videos

Make a photo, an AI actor, or your own video speak a script in any voice, with lip-sync and optional captions.

A talking video makes a face say your words. Start from a photo, one of more than 100 AI actors, or a video of someone speaking, then type a script or upload a recording. The lips move with the audio.

<Frame>
  <img alt="The Talking Videos page with the AI actor Sarah selected, an Add Product (Optional) tile, Generate selected as the audio source, a script for Example Yoga Studio, and a Generate Talking Video (34 credits) button; example videos on the right" />
</Frame>

Open **AI Talking Videos** in the sidebar, or choose **Videos** > **Talking Videos**.

## Make a talking video

<Steps>
  <Step title="Choose who speaks">
    Under **Choose Upload Type**, pick one:

    * **Image**: a photo to bring to life. It must be at least 512 × 512 pixels; one clear, front-facing face works best.
    * **Video**: a video of someone speaking, MP4 or MOV, 3 seconds to 5 minutes. The lips are matched to your new audio, which replaces the original sound.
    * **AI Actor**: pick a ready-made presenter in **Select AI Actor**, or click **Create Actor** to make your own. Picking an actor also picks the actor's voice.

    <Frame>
      <img alt="The Select AI Actor window with a search box, Create Actor button, All Actors 107 and My Actors 0 tabs, and actors such as Sarah - Professional Presenter, Lisa - Marketing Expert, Emma - Lifestyle Vlogger, and Alex - Tech Expert" />
    </Frame>

    With an actor, you can also **Add Product (Optional)**: upload a product photo, at least 512 × 512 pixels, and the actor is redrawn holding it. This uses 7 image & video credits.
  </Step>

  <Step title="Add the words">
    Under **Choose Audio Source**:

    * **Generate**: type the script in **Text to speak**, up to 2,500 characters (about 5 minutes). **Estimated length** shows how long it will be and what it will cost.
    * **Upload**: use your own recording, such as a voiceover or a podcast clip. MP3, WAV, M4A, AAC, or OGG, up to 5 minutes.
  </Step>

  <Step title="Pick a voice">
    If you chose **Generate**, click **Change voice** to choose from hundreds of voices. Search by name, or filter by **Language** and **Accent**, and pick a voice that speaks the language of your script.

    To speak in your own voice, click **Clone Your Voice** and upload or record a sample. Choose an **Instant** clone for talking videos; it works on every plan. How many voices you can clone depends on your plan.
  </Step>

  <Step title="Choose captions, model, and resolution">
    <Frame>
      <img alt="Voice Haley Maven - Social Media Bestie with Change voice and Clone Your Voice, an estimate of 34 credits for about 12 seconds at 3 credits per second with the Premium model at 480p, a Generate captions switch, Avatar Model Premium (3-7 credits/sec) or Standard (1-2 credits/sec), Resolution 480p or 720p HD, and a Generate Talking Video (34 credits) button" />
    </Frame>

    * **Generate captions** adds word-by-word captions. Pick the **Highlight color** for the word being spoken.
    * **Avatar Model**: **Premium** gives the best lip-sync. **Standard** costs less.
    * **Resolution**: **480p** or **720p**. Premium videos made from a video keep the resolution of that video.
  </Step>

  <Step title="Review and generate">
    Click **Generate Talking Video**. With a typed script, **Review before generating** opens first, so you can listen to the voice before you pay for the video. Click **Change text or voice** to go back, or **Confirm & Generate**. With an uploaded recording, the video starts right away.

    <Frame>
      <img alt="The Review before generating window: the actor, an audio player with the 10-second preview, a Change text or voice link, the voice, Premium 480p, about 34 credits, and Cancel and Confirm & Generate buttons" />
    </Frame>
  </Step>
</Steps>

The video appears on the right as **Processing video...** and is usually ready in 2 to 5 minutes. Older talking videos are in **History** > **Videos**.

<Frame>
  <img alt="The details of the finished talking video: 31 credits used, the script as the prompt, the creation date, a Share with Team switch, and buttons for Edit with AI, Upscale video, Extend video, Regenerate, and Download video, next to the vertical video of the actor speaking" />
</Frame>

## Credits

Talking videos are charged per second of speech:

| Avatar model | From an image or actor | From a video |
| - | - | - |
| **Premium** | 3 per second at 480p, 7 at 720p | 3 per second, at the video's own resolution |
| **Standard** | 1 per second at 480p, 2 at 720p | 1 per second at 480p, 3 at 720p |

The price on the button is an estimate from the length of your script. When the video is done, the charge is corrected to its actual length, so it can end up slightly lower or higher. In the example above, the estimate was 34 credits and the 10-second video cost 31.

The voice itself doesn't use your text-to-speech characters. Captions are free.

## On the Free plan

* Only the **Standard** model at **480p** is available.
* A video can be as many seconds long as you have credits left, at 1 credit per second.

## Your own actors

**Create Actor** walks you through four steps: **Photo**, **Voice**, **Details**, and **Preview**. Upload a clear photo of a face (PNG, JPG, or WEBP, at least 512 × 512 pixels), or make one with **Generate with AI** or **Transform with AI**, which use image & video credits. Then pick the actor's voice and give them a name. Your actors appear under **My Actors**, and the window shows how many your plan allows, such as **Custom actors: 0/4**.

## If something goes wrong

* If a video fails, you get its credits back. It either disappears from the list or shows **Generation Failed**.
* **Image must be at least 512x512 pixels**: use a larger photo.
* If the lip-sync looks off, try **Premium**, and a photo where the face is straight on and not covered.

<Columns>
  <Card title="Video Agent" icon="robot" href="/docs/videos/video-agent">
    Build a full video with scenes, voiceover, and music.
  </Card>

  <Card title="Manage your videos" icon="folder-open" href="/docs/videos/manage">
    Download, share, and edit your talking videos.
  </Card>
</Columns>


# Video templates
Source: https://easy-peasy.ai/docs/videos/templates

Start from a ready-made video: pick a template, swap in your own photo, and generate your version.

Templates are finished videos you can remake. Each one comes with its prompt, model, length, and shape already set, and many are built around a photo of a person, so you can put yourself in the scene.

<Frame>
  <img alt="The Templates tab with category chips All 126, New 57, Starring You 47, Movies 9, Anime 3, Cartoons 2, Fantasy 2, Fashion 1, and Portraits 1, and template previews such as Jellyfish Sky Ride, Last Slice Office Brawl, Western Stagecoach Showdown, Giant Dandelion Ride, Pulled Into the Game, and Ocean Sai Dance Battle" />
</Frame>

## Use a template

<Steps>
  <Step title="Open the templates">
    Click the **Templates** tab above your results, or the **Templates** button at the top of the page. Pick a category to narrow the list: **Starring You** has templates made for your own photo, and **New** has the latest additions.
  </Step>

  <Step title="Click a template">
    The template fills in the prompt, the model, the length and shape, and any sample photos it needs. Its name appears at the top of the page, with a button to share the template and **✕** to remove it.

    <Frame>
      <img alt="The Giant Dandelion Ride template applied: its name at the top with share and remove buttons, a sample Character photo under References (optional), the filled-in prompt, 15s, 16:9, 480p, Audio, and Generate (90 credits)" />
    </Frame>
  </Step>

  <Step title="Put yourself in it">
    If the template uses a sample photo, such as the **Character** above, remove it with **✕**, then click **Add media** and add a photo of yourself or your product. Keep the prompt as it is: it already refers to the photo as **@image1**.
  </Step>

  <Step title="Adjust and generate">
    Change anything you like, such as the length or the wording, and click **Generate**. The button shows the price for the template's settings.
  </Step>
</Steps>

Removing the template with **✕** clears the settings it filled in, except the ones you've changed.

## Templates that open other tools

A few templates open a different tool with everything set up: [Motion Control](/docs/videos/motion-control), [talking videos](/docs/videos/talking-videos), or a workflow in a new tab.

## What templates cost

A template costs what its model costs for its settings, shown on the **Generate** button. Many **Starring You** templates make 15-second videos with Seedance 2.5 Reference, which costs 90 credits at 480p. On the Free plan, only the templates that use Seedance 2.0 Mini work.

## Examples

The **Examples** tab shows videos other people chose to share publicly. **Recreate** copies an example's prompt into the form, so you can make your own version.

<Columns>
  <Card title="Use references" icon="user" href="/docs/videos/references">
    How reference photos work.
  </Card>

  <Card title="Generate a video" icon="clapperboard" href="/docs/videos/generate">
    Build a video from scratch.
  </Card>
</Columns>


# Upscale a video
Source: https://easy-peasy.ai/docs/videos/upscale

Make a video sharper and larger, up to 4K, with three upscaling models.

Upscaling makes a video sharper and increases its resolution, up to 4K. The content and length stay the same.

<Note>
  Video upscaling is available on paid plans.
</Note>

## Upscale a video

<Steps>
  <Step title="Choose the video">
    On one of your videos, open the **⋮** menu and click **Upscale video**. Or open **Videos** > **Upscale Video** and click **Upload video**: MP4 or MOV, up to 10 minutes and 100 MB.
  </Step>

  <Step title="Pick a model and resolution">
    Choose **Standard**, **Pro**, or **Ultimate**, and a **Resolution**: **720p**, **1080p**, **2K**, or **4K**. The default is Standard at 1080p.

    <Frame>
      <img alt="Upscale Video with the yoga studio clip under Video to upscale, model cards Standard (2 credits), Pro (10 credits), and Ultimate (8 credits), Resolution 720p, 1080p (selected), 2K, and 4K, the note Billed for 5s · 2 credits, and an Upscale video (2 credits) button" />
    </Frame>
  </Step>

  <Step title="Click Upscale video">
    The upscaled video appears in your results when it's ready. Short clips take a few minutes; long videos can take several hours. You can leave the page and find the result in **History** > **Videos**.
  </Step>
</Steps>

## Models and credits

| Model | What it does | 720p | 1080p | 2K | 4K |
| - | - | - | - | - | - |
| **Standard** | Improves clarity and resolution at a lower cost | 0.25 | 0.25 | 0.5 | 1 |
| **Pro** | Restores finer detail | 2 | 2 | 2.5 | 3.5 |
| **Ultimate** | Recovers fine textures and reduces flicker | 1 | 1.5 | 2.5 | 4 |

Prices are in credits per second of the video. The length is rounded up to whole seconds, with a 3-second minimum, and the total is rounded up. For example, a 1-minute video costs 15 credits with Standard at 1080p and 240 credits with Ultimate at 4K.

## Upgrade a MiniMax H3 video to 2K

Videos made with MiniMax H3 at 768p have one more option: **Upgrade to 2K · MiniMax** in the video's menu. Instead of upscaling the pixels, MiniMax H3 renders the video again at 2K, guided by the original video, prompt, and references. The result is saved as a new video and can differ slightly from the original.

<Frame>
  <img alt="The Upgrade to 2K with MiniMax window: the original video, Video seconds billed 5, Total credits 13, a note that MiniMax bills the original references again, and Cancel and Upgrade to 2K · 13 credits buttons" />
</Frame>

It costs 2.5 credits per second of video, plus 2.5 per second of any reference video and 1.25 for each reference image after the first five. The window shows the total before you confirm.

<Columns>
  <Card title="Extend a video" icon="forward" href="/docs/videos/extend">
    Make the video longer.
  </Card>

  <Card title="Manage your videos" icon="folder-open" href="/docs/videos/manage">
    Download and share the result.
  </Card>
</Columns>


# Video Agent
Source: https://easy-peasy.ai/docs/videos/video-agent

Describe a video in a sentence, and the agent writes the script, makes the scenes, adds a voiceover, music, and captions, and opens the result in the Video Editor.

The Video Agent turns an idea into a complete, edited video. It writes the script, picks or generates visuals for each scene, records a voiceover, adds music and captions, and assembles everything as a project in the [Video Editor](/docs/videos/video-editor), where you can change anything before you export.

Open **Videos** > **Video Agent**. The Video Agent is in beta.

<Frame>
  <img alt="The Video Agent page: Describe your idea. We'll make the video., an idea for a 30-second Instagram reel for Example Yoga Studio, Reference images and Surprise me buttons, Format & length with Portrait 9:16 selected and a 30s target duration, and a Brief panel with Visuals AI image story, Voice Hope, and Music Tourist" />
</Frame>

## Make a video

<Steps>
  <Step title="Describe your idea">
    In **Your idea**, say what the video is for, who it's for, the tone, and what should happen. A sentence works; a paragraph works better. For example: "A 30-second Instagram reel for Example Yoga Studio announcing a new sunrise class every weekday at 6:30. Calm, warm, inviting tone. Show the studio at dawn, a small group stretching, and end with: your first class is free."

    **Surprise me** fills in an example idea. **Reference images** lets you add up to 6 photos, such as a real property, product, or place, and the scenes are based on them.
  </Step>

  <Step title="Choose the format and length">
    Under **Format & length**, pick **Portrait** (9:16, the default), **Landscape** (16:9), or **Square** (1:1), and a **Target duration** from 10 to 200 seconds. **Auto** lets the agent pick the length, usually 20 to 60 seconds.
  </Step>

  <Step title="Choose the visuals">
    <Frame>
      <img alt="The Visuals section: Media type cards Image story (selected), Video clips, Stock videos, and Your own media; Characters (optional) with Create your first character; and Style cards Default (selected), Cinematic, Ghibli Studio, Educational, Pixar, Anime, Watercolor, and Comic Book" />
    </Frame>

    | Media type | What the scenes are made of |
    | - | - |
    | **Image story** (default) | An AI image for each scene, with motion, timed to the narration. |
    | **Video clips** | An AI video clip for each scene. Pick the **Video model**; each shows its price per second. Paid plans only. |
    | **Stock videos** | Real stock footage that matches your script. |
    | **Your own media** | Up to 20 of your own photos and videos, 50 MB each, with narration written around them. **Enhance photos** improves them with AI, and **Turn images into videos** animates each photo (paid plans). |

    **Characters (optional)** puts the same person, with the same face, in every scene: pick one of your [custom actors](/docs/videos/talking-videos#your-own-actors) or create one. **Style** sets the look, from **Cinematic** to **Anime** or **Watercolor**; **Show all 27** lists them all.
  </Step>

  <Step title="Choose the voice and music">
    <Frame>
      <img alt="The Audio section: Voice with No voiceover or Hope (American) with Change and a play button; Background music tabs Library (selected), Auto, AI Generate, and No music; category chips; and a list of tracks with Tourist selected" />
    </Frame>

    * **Voice**: the narrator. Click **Change** to pick another voice, or choose **No voiceover**.
    * **Background music**: pick a track from the **Library**, let the agent choose (**Auto**), describe a track to create (**AI Generate**, always instrumental), or choose **No music**.
  </Step>

  <Step title="Check the advanced settings (optional)">
    Open **Advanced** to set:

    * **Language** of the narration and captions. **Auto** follows your idea.
    * **Image model** for image scenes. The default is **GPT Image 2.5 Flare Medium**.
    * **Captions** on or off, and the **Caption style**: **Viral** (the default), **Hormozi**, **Wrap**, **Faceless**, **Classic**, or **Minimal**. You can change the style later in the editor.
  </Step>

  <Step title="Click Generate video">
    The agent starts working and takes you to the **AI Video Editor**, where your project shows **Generating** with a timer. It usually takes 2 to 5 minutes, longer for long videos, and you can leave the page in the meantime.

    <Frame>
      <img alt="The AI Video Editor page with a New Project button and, under Recent Projects, the finished 9:16 project A 30-second Instagram reel for Exa..." />
    </Frame>
  </Step>
</Steps>

When it's done, click the project to open it. The video is laid out on tracks for the scenes, on-screen text, music, narration, and captions. Watch it, change what you want, and **Export** it as an MP4. See [Video Editor](/docs/videos/video-editor).

## Credits

There's no total price before you start. Each part is charged as it's made:

| Part | Cost |
| - | - |
| Scene images | The image model's price per image, 3 credits each with GPT Image 2.5 Flare Medium |
| Video clips | The video model's price per second, shown on each model. Clips without sound cost 20% less. |
| Stock videos and your own photos | Free |
| **Enhance photos** | 4 credits per photo |
| Voiceover | Text-to-speech characters: half the length of the narration |
| Music | **Library** tracks are free. **Auto** and **AI Generate** use 600 text-to-speech characters. |
| Script, captions, and planning | Free |
| Exporting the MP4 in the editor | 2 credits |

On paid plans, the agent checks your balance first and stops with a message if the video would need more credits than you have left. Try a shorter length or a cheaper video model.

If a project fails, it's marked **Generation failed**. Scenes that were already made are still charged; scenes that failed are not.

## On the Free plan

* **Image story**, **Stock videos**, and **Your own media** are available; **Video clips**, characters, and **Turn images into videos** need a paid plan.
* Background music comes from the **Library**.
* You can make 2 projects per billing period, and you need at least 1 credit left.

## Tips

* Say who the video is for and what the viewer should do at the end, such as "book a free class".
* Put names, prices, and times in your idea exactly as they should appear.
* If a scene isn't right, don't start over: open the project and [regenerate that scene](/docs/videos/video-editor#regenerate-a-clip-with-ai).

<Columns>
  <Card title="Video Editor" icon="film" href="/docs/videos/video-editor">
    Edit and export the project.
  </Card>

  <Card title="Talking videos" icon="comment" href="/docs/videos/talking-videos">
    A single presenter speaking to camera.
  </Card>
</Columns>


# Video Editor
Source: https://easy-peasy.ai/docs/videos/video-editor

Arrange video, images, voiceover, music, text, and captions on a timeline, regenerate any clip with AI, and export an MP4.

The Video Editor is where [Video Agent](/docs/videos/video-agent) projects open, and where you can build a video yourself from your clips, images, and audio. Every change saves automatically. The editor is in beta.

<Frame>
  <img alt="The Video Editor: the media library with images, voiceover clips, and music on the left; the vertical preview with the text New Sunrise Yoga and a caption in the middle; and a timeline with Visual scenes, Scene text, Music, Narration, and Captions tracks at the bottom; Captions, Text, Settings, and Export buttons at the top" />
</Frame>

## Open or start a project

Open **Videos** > **Video Editor**. Your projects are under **Recent Projects**; click one to open it.

* **New Project** starts an empty project in 16:9.
* A project's menu has **Duplicate**, **Share with Team** (on Teams plans), and **Delete**. Deleting a project can't be undone.

To rename a project, click its title at the top. **Settings** changes the title and the **Aspect Ratio**: **16:9 Landscape**, **9:16 Portrait**, or **1:1 Square**.

## Add media

The **Media** panel on the left holds everything in the project.

| Button | What it does |
| - | - |
| **Import** | Add your past videos, images, music, voiceovers, and sound effects from **Media Gallery**. Search by prompt and click **Add to Timeline**. |
| **Upload** | Add video, image, and audio files from your computer. |
| **Generate** | Open AI Images, AI Videos, talking videos, music, sound effects, or text to speech in a new tab to make something new, then import it. |

Click **Add** on a media card, or double-click it, to put it on the timeline.

## Work on the timeline

* **Move** a clip by dragging it. **Trim** it by dragging either edge. Clips on the same track can't overlap.
* **Play**, pause, and jump 5 seconds back or forward with the controls above the tracks.
* **Delete** removes the selected clip. You can also press Delete or Backspace.
* **Undo** and **Redo**: Ctrl+Z and Ctrl+Shift+Z (⌘ on a Mac).
* **Zoom** the timeline from 25% to 400%.
* Each track can be muted, hidden, or locked, and has its own volume.

## Change a clip

Click a clip to open **Clip Properties**.

| Clip | What you can change |
| - | - |
| Video | **Volume**, **Fade In**, **Fade Out** |
| Image | **Animation Presets**, **Motion**, how it comes **In** and goes **Out**, and **Fit** (Fill or Fit) |
| Audio | **Volume**, **Fade In**, **Fade Out** |
| Text | The text, **Font**, **Size**, **Weight**, **Color**, **Position**, **Align**, **Offset**, and **Background** |

**Text** at the top adds a new text clip.

## Regenerate a clip with AI

Not happy with one scene? Select it and use **Regenerate with AI** at the top of **Clip Properties**. You don't have to remake the whole video.

<Frame>
  <img alt="Clip Properties for an image scene: Regenerate with AI with the scene's prompt, the reference image of the character, Mode With reference or New image, the model GPT Image 2.5 Flare Medium, and a Regenerate button" />
</Frame>

| Clip | How to regenerate it | Cost |
| - | - | - |
| Image | Edit the prompt, choose **With reference** to keep the same person or **New image**, and pick a model. | The image model's price |
| Video | Choose **Re-animate frame** or **New video**, pick a model, and set a length from 2 to 15 seconds. Paid plans only. | Shown on **Regenerate · N credits** |
| Voiceover | Edit the **Script** and pick a **Voice**. The timeline and captions adjust to the new length. | Text-to-speech characters: half the length of the script |
| Music | Describe the music in **Music description**. | 600 text-to-speech characters |
| Sound effect | Describe the sound in **Sound description**. | 1,000 text-to-speech characters |

## Captions

Click **Captions** to caption your video. **Generate Captions** transcribes the voiceover and video clips you select, in the language you choose or with **Auto-detect**. Captions are free, but you need at least 1 credit left to make them.

Click a caption clip to style it in **Caption Properties**: the style, position, font size, how many words show at once (**Words per page**), and the **Highlight Color** of the word being spoken. **Apply style to all captions** copies the style to every caption. To fix words or timing, click **Edit caption text**: edit each line and its start and end times, merge or split lines, or use **Find** and **Replace with**.

## Export your video

<Steps>
  <Step title="Click Export">
    **Export Video** shows the project details. Choose **1080p (Full HD)** or **4K (Ultra HD)**.

    <Frame>
      <img alt="The Export Video window with Project Details (title, aspect ratio 9:16, 9 clips, 1 video and image track, 2 audio tracks), Resolution 1080p (Full HD) or 4K (Ultra HD), and an Export Video button" />
    </Frame>
  </Step>

  <Step title="Click Export Video">
    Rendering takes a few minutes and shows its progress. Keep the editor open until it finishes; you can close the export window itself.
  </Step>

  <Step title="Download">
    When you see **Export complete!**, click **Download Video**. The MP4 is also saved in **History** > **Videos**.

    <Frame>
      <img alt="The Export Video window after rendering: Export complete!, a Download Video button, and Export again" />
    </Frame>
  </Step>
</Steps>

Each export costs 2 image & video credits. Videos are exported as MP4 at 30 frames per second: 1920 × 1080 for landscape, 1080 × 1920 for portrait, and 1080 × 1080 for square, or twice that in 4K.

<Columns>
  <Card title="Video Agent" icon="robot" href="/docs/videos/video-agent">
    Start a project from one sentence.
  </Card>

  <Card title="Manage your videos" icon="folder-open" href="/docs/videos/manage">
    Find, download, and share your exports.
  </Card>
</Columns>


# Build and run a workflow
Source: https://easy-peasy.ai/docs/workflows/build

Add nodes to the canvas, connect them, pass results between them with @ references, and run them.

This page builds a two-step workflow: a text description goes into an image model, which makes a promo image for a yoga class.

<Frame>
  <img alt="The workflow editor for Class promo image: a #1 Text Input node describing a bright yoga studio at dawn, connected to a #2 Text to Image node with Model Nano Banana Pro, Size 1:1, the prompt @#1 Text Input, Number of images 1, and Run (7 credits), with Add Node and zoom controls" />
</Frame>

## Build a workflow

<Steps>
  <Step title="Add your first node">
    Click **Add Node** at the top right and pick one, such as **Text Input**. It appears in the middle of the canvas.

    <Frame>
      <img alt="The Add Node list: Text Input, Upload Media, Select Media, Text to Image, Image to Image, Upscale Image, Image Splitter, Text to Video, and Image to Video" />
    </Frame>
  </Step>

  <Step title="Connect the next node">
    Drag from the round handle on the right of a node to the left handle of another. Or drop the connection on an empty spot: a list of nodes that can take that output opens, with a search box.

    <Frame>
      <img alt="The Add text node list that opens after dropping a connection, with Search nodes and the options Text to Image, Text Generation, Text to Video, Text to Speech, Output, and First-Last Frame to Video" />
    </Frame>

    Handles are colored by type: blue for text, purple for images, cyan for video, and indigo for audio. You can only connect matching types.
  </Step>

  <Step title="Use the result in the prompt">
    Connecting into a prompt adds a reference like `@#1 Text Input`. When the workflow runs, the reference is replaced by that node's result. Type `@` in any prompt to reference another connected node, and add your own words around it, for example: "`@#1 Text Input`, in the style of a watercolor painting".
  </Step>

  <Step title="Set up each node">
    Pick the model and settings in each node. The **Run** button shows what a node costs before you run it.
  </Step>
</Steps>

## Run a workflow

Click **Run** in a node, or the ▶ in its header. Before it runs, any node it depends on that doesn't have a result yet runs first, so running the last node runs the whole chain. Nodes that don't depend on each other run at the same time.

<Frame>
  <img alt="The same workflow after running: the #2 Text to Image node shows the generated image of a small group walking into a bright yoga studio with plants and wooden floors, with a green check" />
</Frame>

* A running node has a blue border. A finished one gets a green check, and its settings fold away; click the node's settings icon to show them again.
* A node that fails turns red and shows why, for example a missing prompt or a model that isn't on your plan.
* Results stay in the node. Open or download them from the node's menu. Images and videos you make also appear in [AI Images](/docs/ai-images/manage) and [AI Videos](/docs/videos/manage), and in History.

**To make a new version**, change the input and click **Regenerate** on the node you want to redo. Nodes that already have a result are reused, not run again, so regenerate each node you want to change.

<Warning>
  Workflows run in your browser. Keep the page open until your nodes finish; if you leave, the page warns you, and nodes that were still running stop updating.
</Warning>

## Nodes

<AccordionGroup>
  <Accordion title="Inputs">
    | Node | What it does |
    | - | - |
    | **Text Input** | Text you type, such as a topic, a script, or a prompt |
    | **Upload Media** | An image (up to 10 MB), video (up to 50 MB), or audio file (up to 20 MB) from your computer. Upload several files and each gets its own node |
    | **Select Media** | An image or video you made before in AI Images or AI Videos |
  </Accordion>

  <Accordion title="Images">
    | Node | What it does |
    | - | - |
    | **Text to Image** | Makes 1 to 4 images from a prompt, with any [image model](/docs/ai-images/models). Extra images each get their own node |
    | **Image to Image** | Edits connected images with an instruction, like [editing in AI Images](/docs/ai-images/edit) |
    | **Upscale Image** | Makes an image 2x or 4x larger, in **Crisp**, **Creative**, or **Restore old photo** mode |
    | **Image Splitter** | Cuts an image into a grid of up to 5 by 5 pieces, each in its own node |
  </Accordion>

  <Accordion title="Videos">
    | Node | What it does |
    | - | - |
    | **Text to Video** | Makes a video from a prompt |
    | **Image to Video** | Animates an image, with an optional prompt |
    | **First-Last Frame to Video** | Makes a video that moves from a first frame to a last frame |
    | **References to Video** | Makes a video that keeps the characters or products from reference images |
    | **Merge Videos** | Joins connected videos into one, in the order you set |
    | **Talking Video** | Makes a person in an image or video speak a script or an audio file, like [talking videos](/docs/videos/talking-videos) |

    Video nodes have the same models, lengths, aspect ratios, and resolutions as [AI Videos](/docs/videos/models). Their outputs include the video's first and last frames, to chain clips together.
  </Accordion>

  <Accordion title="Text and audio">
    | Node | What it does |
    | - | - |
    | **Text Generation** | Writes text from a prompt, and can read a connected image. Choose the model, and add an optional system prompt |
    | **Text Preset** | Runs one of the [tools](/docs/tools/overview), with its own fields |
    | **Text to Speech** | Reads connected or typed text aloud in a voice you pick |
    | **Music Generation** | Makes a song or instrumental from a description, or from your style and lyrics |
  </Accordion>

  <Accordion title="Organize">
    | Node | What it does |
    | - | - |
    | **Note** | A sticky note for instructions or comments. It doesn't run |
    | **Output** | Shows a final result. Add it by dropping a connection on an empty spot |

    Select several nodes and click **Group** (or press Cmd/Ctrl + G) to put them in a group you can color, lock, and move together.
  </Accordion>
</AccordionGroup>

<Tip>
  In **Text Generation**, **Text to Speech**, and **Talking Video**, choose the model or voice yourself before running, rather than relying on the one shown.
</Tip>

## Work on the canvas

| To | Do this |
| - | - |
| Move around | Scroll, or drag the canvas in pan mode (the hand icon) |
| Select several nodes | Switch to select mode (the arrow icon) and drag a box |
| Zoom | Pinch, the + and − buttons, or Cmd/Ctrl + = and Cmd/Ctrl + − |
| Fit everything on screen | Cmd/Ctrl + 0 |
| Undo or redo | Cmd/Ctrl + Z, and Cmd/Ctrl + Shift + Z |
| Rename a node | Double-click its title |
| Copy or delete a node | Use the buttons, or the ⋯ menu, on the node |

The top bar shows **All changes saved** once your edits are saved. The download button exports the workflow as a file.

## What each node costs

Each node uses the same allowance, at the same price, as the feature it runs.

| Nodes | Use |
| - | - |
| Text to Image, Image to Image, Upscale Image, video nodes, Talking Video | Image & video credits, shown on the **Run** button. Upscaling costs 2 credits at 2x and 4 at 4x |
| Text Generation, Text Preset | Words, like [Marky](/docs/marky/models) and [tools](/docs/tools/models) |
| Text to Speech | Text-to-speech characters, one per character |
| Music Generation | 600 text-to-speech characters per song |
| Text Input, Upload Media, Select Media, Image Splitter, Merge Videos, Note, Output | Free |

Whoever runs a node pays for it, including in a [shared workflow](/docs/workflows/share).

On the Free plan, a node set to a premium model fails with a message to upgrade. Pick a model that's available on Free; the default image model, Nano Banana Pro, isn't. See [Plans and credits](/docs/account/plans-and-credits#the-free-plan).

<Columns>
  <Card title="Share and collaborate" icon="users" href="/docs/workflows/share">
    Invite people and edit together.
  </Card>

  <Card title="Workflows" icon="diagram-project" href="/docs/workflows/overview">
    Templates, import, and export.
  </Card>
</Columns>


# Workflows
Source: https://easy-peasy.ai/docs/workflows/overview

Chain AI steps on a visual canvas, from text to images, videos, speech, and music, and reuse the whole chain with new inputs.

**Workflows** let you connect AI steps on a canvas, so the result of one step feeds the next. Write a script, turn it into images, animate them, add a voiceover, and merge the clips, then change the first step and run it all again.

Each step is a **node**: a text input, an image or video model, a tool, text to speech, and more. You build the chain once and reuse it whenever you need the same kind of result.

<Note>
  Workflows are in beta, and the editor works on desktop screens only.
</Note>

## Open Workflows

Click **Workflows** in the sidebar, or find it in **All Tools** > **AI Agents**.

<Frame>
  <img alt="AI Workflow Builder, Design powerful, automated multi-step AI workflows for images, video, and more, with Create Workflow, Browse Templates, and Import Workflow buttons, How to Start, and the My Workflows and Templates tabs" />
</Frame>

* **My Workflows** lists your workflows. Search finds them by name.
* **Shared with me** lists workflows other people shared with you. It appears once someone does.
* **Templates** has ready-made workflows for images and videos, such as virtual try-on, virtual staging, a product commercial, and an animated short film.

**How to Start** plays a short video tour.

## Create a workflow

<Tabs>
  <Tab title="From scratch">
    Click **Create Workflow**. An empty canvas opens, named with today's date. Click the name at the top to rename it. See [Build and run a workflow](/docs/workflows/build).
  </Tab>

  <Tab title="From a template">
    Click **Browse Templates**, then a template to preview it. Click **Use Template** at the top right to copy it into your workflows, then change the inputs and run it.
  </Tab>

  <Tab title="From a file">
    Click **Import Workflow** and choose a workflow `.json` file exported from Easy-Peasy.AI.
  </Tab>
</Tabs>

## Manage your workflows

Click the ⋯ menu on a workflow's card:

<Frame>
  <img alt="A workflow card, Class promo image, with its menu open: Edit, Duplicate, Export, Add to favorites, Upgrade to Teams, and Delete" />
</Frame>

| Option | What it does |
| - | - |
| **Edit** | Opens the workflow |
| **Duplicate** | Makes a copy, including the results already in its nodes |
| **Export** | Downloads the workflow as a `.json` file, to back it up or import it into another account |
| **Share with Team** | On a Teams plan, shares the workflow with your team. On other plans this reads **Upgrade to Teams** |
| **Delete** | Deletes the workflow. Only its owner can |

Workflows save automatically as you edit.

## What workflows cost

Workflows have no charge of their own, and there's no limit on how many you make. Each node uses the same allowance as the feature it runs: image and video nodes use image & video credits, text nodes use words, and speech and music use text-to-speech characters. Moving data between nodes, splitting images, and merging videos are free. See [Build and run a workflow](/docs/workflows/build#what-each-node-costs).

<Columns>
  <Card title="Build and run a workflow" icon="diagram-project" href="/docs/workflows/build">
    Add nodes, connect them, and run them.
  </Card>

  <Card title="Share and collaborate" icon="users" href="/docs/workflows/share">
    Work on a workflow with other people.
  </Card>
</Columns>


# Share and collaborate
Source: https://easy-peasy.ai/docs/workflows/share

Invite people to view or edit a workflow, share it by link or with your team, and work on it together in real time.

Share a workflow to build it with a colleague, hand a finished one to a client, or give your team a set of ready-made workflows.

## Invite people

Open the workflow and click the people icon at the top right. The **Share workflow** window opens.

<Frame>
  <img alt="The Share workflow window: Add people by email with the role Editor and an Invite button, People with access listing the owner, and General access with Anyone with the link off and Your team available on the Teams plan with an Upgrade link" />
</Frame>

<Steps>
  <Step title="Add their email address">
    Type it in **Add people by email**. They need an Easy-Peasy.AI account; if they don't have one, ask them to sign up first.
  </Step>

  <Step title="Choose a role">
    **Editor** can change the workflow and run it. **Viewer** can only look at it.
  </Step>

  <Step title="Click Invite">
    They get an email, and the workflow appears in their **Shared with me** tab. To remove someone, use the remove button next to their name under **People with access**.
  </Step>
</Steps>

## Share by link or with your team

Under **General access**:

* **Anyone with the link**: turn it on, choose **Viewer** or **Editor**, and click **Copy link**. People who open the link sign in and are added with that role.
* **Your team**: on a Teams plan, turn it on so your team members can view and edit the workflow. You can also do this with **Share with Team** in the workflow card's menu.

Turn a switch off to stop sharing that way. People who already opened the link stay under **People with access** until you remove them.

## Work together

People with the workflow open see each other's cursors, and the share window shows how many are active now. Changes save automatically.

Only the owner can change who has access, and delete the workflow.

<Note>
  Whoever runs a node pays for it from their own plan. A viewer can't run nodes.
</Note>

<Columns>
  <Card title="Build and run a workflow" icon="diagram-project" href="/docs/workflows/build">
    Nodes, connections, and running.
  </Card>

  <Card title="Teams" icon="users" href="/docs/account/team">
    Add teammates to your plan.
  </Card>
</Columns>