OpenAI released o3-pro, an upgraded reasoning model available to ChatGPT Pro, Team, Enterprise, and Edu users as well as via developer API at $20 per million input tokens and $80 per million output tokens. The model performs step-by-step reasoning and achieves higher benchmark scores than competing frontier models in math (AIME 2024) and PhD-level science knowledge (GPQA Diamond) evaluations. Expert reviewers prefer o3-pro over o3 in all tested categories and rate it higher for clarity, comprehensiveness, instruction-following, and accuracy. For financial institutions tracking frontier model capability progression and pricing strategy, o3-pro represents another generation of benchmark gains and API-tier differentiation among hyperscaler frontier offerings.
179 other tracked developments touching the same claims.