On 31 January 2025, OpenAI introduced the o3-mini model, marking a significant advancement in the company’s cost-efficient artificial intelligence development. Replacing the previous o1-mini, the new model features enhanced STEM capabilities while delivering a 24% faster response time than its predecessor.
The o3-mini delivers outstanding performance in mathematics, coding, and scientific tasks. In external testing, 56% of users preferred its responses over those of the o1-mini, and it made 39% fewer critical errors when solving complex, real-world problems. The model operates at three distinct reasoning levels—low, medium, and high—allowing users to adjust computational resources based on task complexity. Additionally, the o3-mini supports long-requested developer features such as function calls, structured outputs, and developer messages, making it production-ready from the outset.
The o3-mini comes with significantly more competitive pricing—63% cheaper than the o1-mini and 93% less expensive than the o1 model—at a rate of $1.10/$4.40 per million tokens (input/output), with a 50% cache discount. ChatGPT Free, Plus, Team, and Pro users already have access to the o3-mini, while enterprise users will gain access in February. Plus and Team users have seen their daily message limits raised from 50 to 150, while Pro users now enjoy unlimited access to both the o3-mini and the o3-mini-high variant. Additionally, the o3-mini boasts a 200,000-token context window, surpassing the 128,000-token capacity of its competitor, DeepSeek-R1.
Sources:
Discover the latest reasoning model from OpenAI with improved performance.
2.
Carl Franzen
