OpenAI has released o3-pro, a new reasoning model built for more complex tasks and longer inputs. The company says the system can handle up to 200,000 tokens of context, giving it far more room to process extended documents, code, and multi-step instructions.

Alongside the larger context window, OpenAI says o3-pro includes updated safety measures intended to reduce hallucinations and improve reliability in demanding use cases. The launch reflects the company’s push to make its reasoning models more useful for enterprise workflows, research, and other settings where accuracy matters.

Reasoning-focused models have become a central battleground in the AI sector as companies compete on depth, consistency, and practical performance. By expanding context capacity and adding guardrails, OpenAI is positioning o3-pro as a tool for users who need more than short-form chat responses.

The release adds to a fast-moving race among AI developers to build models that can think through longer problems while staying more dependable. As with other frontier systems, real-world performance will depend on how well the model handles complexity without drifting into errors or unsupported claims.