OpenAI Launches o3-pro Reasoning Model with Lower API Costs

OpenAI released o3-pro on June 10, 2025, as its most advanced reasoning AI model to date. The model replaces o1-pro for ChatGPT Pro and Team users, with access for Enterprise and Education users scheduled for the following week.

Designed for complex tasks in science, mathematics, programming, and writing, o3-pro integrates tools for web search, file analysis, Python execution, and visual reasoning. The company reports that o3-pro outperformed its non-pro counterpart in internal evaluations and surpassed competitors, including Google's Gemini 2.5 Pro on the AIME 2024 benchmark and Anthropic's Claude 4 Opus on GPQA Diamond.

The model is available via API at $20 per million input tokens and $80 per million output tokens, which represents a significant price reduction over its predecessor. Despite these advancements, current limitations include slower response times compared to o1-pro and a lack of support for image generation and the Canvas workspace.