OpenAI is rolling out its latest o3-mini reasoning model inside ChatGPT and its API services, making a version with rate limits available to free ChatGPT users for the first time.
Originally announced during OpenAI’s 12-day “ship-mas” event in December, o3-mini is designed to match the performance of o1 in math, coding, and science while delivering responses 24% faster. Unlike its predecessor, the o3-mini also provides more accurate answers and shows its reasoning step-by-step, rather than just giving a final response.
OpenAI has shared early benchmarks showing o3 models outperforming o1 in various coding and reasoning tasks, and now the o3-mini version promises even better results with reduced costs and latency. Developers can access o3-mini through OpenAI’s API services, including the Chat Completions API, Assistants API, and Batch API.
Paid users will also have access to o3-mini-high, which OpenAI describes as the “best coding option in ChatGPT,” offering higher intelligence responses with slightly longer generation times. Additionally, o3-mini will integrate with search to provide answers that include web source links.
This marks the first time free ChatGPT users can try out OpenAI’s reasoning models, coming just days after Microsoft made o1 free for all Copilot users and DeepSeek made waves in the AI industry. Free users can experiment with o3-mini by selecting the “Reason” feature in the chat bar, with rate limits similar to GPT-4o.
Meanwhile, OpenAI is significantly increasing message limits for Plus and Team users, tripling them to 150 messages per day. Pro users, who pay $200 per month, will have unlimited access to o3-mini. The update rolls out today for ChatGPT Plus, Team, and Pro users worldwide.
