OpenAI o1 ranks in the 89th percentile on competitive programming questions (Codeforces), places among the top 500 students in the US in a qualifier for the USA Math Olympiad (AIME), and exceeds human PhD-level accuracy on a benchmark of physics, biology, and chemistry problems (GPQA). While the work needed to make this new model as easy to use as current models is still ongoing, we are releasing an early version of this model, OpenAI o1-preview, for immediate use in ChatGPT and to trusted API users(opens in a new window).
Our large-scale reinforcement learning algorithm teaches the model how to think productively using its chain of thought in a highly data-efficient training process. We have found that the performance of o1 consistently improves with more reinforcement learning (train-time compute) and with more time spent thinking (test-time compute). The constraints on scaling this approach differ substantially from those of LLM pretraining, and we are continuing to investigate them.
OpenAI introduces its advanced model, OpenAI o1, which showcases exceptional capabilities, ranking in the 89th percentile in Codeforces competitive programming and among the top 500 for the USA Math Olympiad. Additionally, it demonstrates superior accuracy in solving complex science problems, surpassing PhD-level benchmarks in physics, biology, and chemistry. Although further development is required for ease of use, an early version, OpenAI o1-preview, is now available for use in ChatGPT and to select API users. The model benefits significantly from a large-scale reinforcement learning algorithm that enhances its problem-solving efficiency and effectiveness through iterative training and extended processing time. Current research continues to explore and refine the scalable aspects of this technology.
