OpenAI Unveils o1-Pro, Its Most Powerful Reasoning Model to Date

OpenAI has launched o1-pro, a new reasoning model it says is its most capable yet, with availability beginning December 5, 2024 through ChatGPT Pro and the API. The company says the model is designed to handle complex tasks that require deeper step-by-step reasoning, and early results suggest strong performance on demanding benchmarks in math, coding, and advanced science problems.

According to OpenAI, o1-pro outperforms earlier versions on difficult evaluations, including PhD-level science questions, positioning it as a major step forward in AI reasoning. The model is aimed at users who need higher reliability on complex analytical work, from technical problem-solving to research assistance.

But the launch also highlights the growing cost of frontier AI systems. OpenAI has priced o1-pro at $200 per million input tokens, a level that underscores both its computational intensity and the premium nature of the offering. That pricing may limit adoption to enterprise users, developers, and professionals with specialized workloads.

The release reflects a broader industry trend: as models become more capable, they also become more expensive to run. For OpenAI, o1-pro represents a push to lead in advanced reasoning, even as questions remain about affordability, efficiency, and the practical trade-offs of deploying such compute-heavy systems at scale.