OpenAI Launches o3-pro Model for Tackling Complex Challenges: Is It Right for You?

OpenAI has launched o3-pro, an enhanced AI model on ChatGPT, designed to solve complex problems with improved reasoning and accuracy. This upgrade, which succeeded o1-pro, underwent a rigorous “4/4 reliability” evaluation to ensure consistent performance. Although it shares the core architecture with the earlier o3 model, o3-pro operates in “high-compute mode,” requiring more computational power for better results. It excels in areas such as science, math, programming, data analysis, and academic writing, scoring 93% on the AIME 2024 math exam and 84% on the GPQA Diamond benchmark. Users prefer o3-pro’s responses over the standard o3 model. Currently, it is available to ChatGPT Pro and Team users, replacing o1-pro as of June 10, 2025, with enterprise and educational access coming soon. The model can also be accessed via the OpenAI API, with pricing starting at $20 per million input tokens.

Source link