OpenAI Unveils o1 Pro, Boosting AI Reasoning on Math and Coding Benchmarks
Read the full article for context, quotes, and updates from the team.
OpenAI has released o1 Pro, a new reasoning-focused artificial intelligence model that the company says delivers substantial gains over earlier versions on demanding tasks. The model is designed to handle complex problem-solving more effectively, with particularly strong results in mathematics and coding evaluations.
According to OpenAI, o1 Pro achieved top scores on several challenging benchmarks, including an 83% result on the AIME 2024 mathematics competition, a widely watched test of advanced problem-solving ability. The company says the model also performs strongly on coding tasks, underscoring continued progress in AI systems that can reason through multi-step problems rather than simply generate fluent text.
The launch comes as developers and researchers race to improve model reliability and performance on tasks that require deeper analytical thinking. Reasoning models such as o1 Pro are increasingly seen as a key step toward more capable AI tools for software development, scientific work, and technical analysis.
OpenAI said o1 Pro is now available to Pro subscribers, expanding access to its latest high-end model for users seeking stronger performance on complex workloads. The release adds to growing competition in the AI industry as companies push to build systems that can solve harder problems with greater accuracy and consistency.
