OpenAI o1-preview

12 September 2024

The first reasoning model. Thinks before it answers, and spends more compute on harder problems.