News · September 12, 2024
OpenAI previews o1, models that spend more time reasoning
OpenAI has released o1-preview and o1-mini, models trained to work through problems before answering. They target math, science and coding tasks, at the cost of slower responses and fewer features.
What happened
- OpenAI introduced the o1 series on September 12, 2024 and said the models are trained to spend more time thinking before they respond, according to its announcement.
- According to OpenAI, the reasoning model solved 83% of problems on a qualifying exam for the International Mathematics Olympiad, versus 13% for GPT-4o, and reached the 89th percentile in Codeforces coding contests.
- OpenAI said ChatGPT Plus and Team users get access with weekly message limits, API access starts with usage tier 5 developers at 20 requests per minute, and o1-mini is 80% cheaper than o1-preview.
- Fortune, in an article from Bloomberg, reported that the preview lacks features such as web browsing and file and image uploads, and that OpenAI described it as an early model.
Why it matters
This is a different kind of model, not simply a better chatbot. It looks useful for hard, multi-step problems such as tricky bugs, algorithm design or data analysis, and less useful for fast, high-volume tasks where latency and cost matter more than depth.
Our suggestion: keep the current model for everyday product features and test o1 on a small set of difficult internal tasks, such as code review of complex changes or test generation. Track time saved against higher cost and slower answers before building it into a product.
Sources
This post is CMM’s own summary and commentary on publicly reported news. Facts are attributed to the sources linked above; please read them for full details. CMM is not affiliated with these publishers or with the companies mentioned, and trademarks belong to their owners. Prepared with AI assistance. Corrections or requests: [email protected]. See our editorial policy.
Want help with this?
Tell us what you need, and we will suggest the right team model on a short call.
Talk to us