o1 Series

A series of reasoning-focused models from OpenAI whose central feature is deliberative, slow-thinking inference. By internalizing chain-of-thought reasoning as a training objective, the series achieves substantially higher accuracy on mathematics, programming, and complex logical tasks.