OpenAI unveiled its new o3-mini model on Friday, offering a “reasoning” model to all ChatGPT users, including those on the free tier for the first time. The new o3 mini model can be accessed by clicking the “Reason” button in the message interface.
OpenAI’s internal testing shows that the o3-mini processes responses 24% faster than the o1-mini model. According to OpenAI, when configured to “medium” reasoning effort, o3-mini achieves comparable results to the more premium o1 system across various technical fields, including mathematics, programming, and science.
Early benchmarking shows particularly strong results in specialized technical fields, with the model’s performance nearly matching that of its more expensive alternatives when dealing with complex computational and analytical tasks.
OpenAI has integrated transparent reasoning processes into its latest o3-mini model, allowing users to track how the system develops its responses rather than simply receiving finished answers like the regular ChatGPT. The new model comes equipped with built-in web search through ChatGPT Search a feature OpenAI plans to roll out across its entire suite of reasoning models.
o3-mini takes the place of o1-mini in the platform’s selection menu. OpenAI has also announced a 3x increase in daily message limits for ChatGPT Plus and Team subscribers, jumping from 50 to 150 messages per day. For those seeking unlimited access to the new system, OpenAI’s premium Pro tier remains available at $200 monthly.
This development comes soon after DeepSeek unveiled its latest AI model, which took the tech industry by storm with its aggressive pricing and open-source system.