If you have a use case in STEM, coding, and problem-solving, this model brings powerful upgrades while maintaining low latency and cost.
🔥 New Features & Improvements
✅ Optimized for STEM – Stronger in math, coding, and science, delivering more accurate and clearer answers.
✅ Three reasoning effort options – Choose low, medium, or high to balance speed and complexity.
✅ Search-enabled responses – Now finds up-to-date answers with links to relevant sources.
✅ Higher message limits – ChatGPT Plus & Team users now get 3x more messages per day (150 instead of 50).
✅ Production-ready developer tools – Supports function calling, structured outputs, and developer messages.
✅ Free user access – For the first time, free users can try a reasoning model by selecting ‘Reason’ in ChatGPT.
🛠 API Availability & Developer Options
🔹 Available in: Chat Completions API, Assistants API, and Batch API
🔹 Rolling out to API usage tiers 3-5
🔹 Enterprise access coming in February
🔹 o3-mini replaces o1-mini, offering higher rate limits & better reasoning
⚠️ Limitations
❌ No vision support – Unlike OpenAI o1, o3-mini does not handle visual reasoning.
🚀 Why This Matters
o3-mini delivers o1-level intelligence with faster response times, this was a big issue in o1-mini which this model solves.
Release note