Archive — US — techcrunch
OpenAI releases o3-pro, a souped-up version of its o3 AI reasoning model
OpenAI has unveiled o3-pro, its newest AI reasoning model designed for enhanced reliability in fields like physics, math, and coding. Available for ChatGPT Pro and Team users, it costs $20 per million input tokens and $80 for outputs. O3-pro offers superior performance over o3 in evaluations and supports advanced features, though it has some current limitations, including slower response times.
Key points
- OpenAI launched o3-pro, a new reasoning AI model, replacing o1-pro.
- o3-pro is available for ChatGPT Pro and Team users from June 10, 2025, with Enterprise and Edu access following a week later.
- Priced at $20 per million input tokens and $80 for output tokens, making it commercially viable.
- Outperforms o3 in expert evaluations in science, education, programming, and writing.
- Introduces advanced tools for web searching, file analysis, and visual reasoning but has slower response times.
- Current limitations include disabled temporary chats and lack of image generation.
Why it matters
- Reflects the growing trend of specialized AI models focusing on reasoning and step-by-step problem-solving.
- Aligns with the 2025 IT trend towards enhancing AI capabilities in education and programming support.
- Demonstrates the competitive landscape in AI, particularly against models like Google’s Gemini and Anthropic’s Claude.
OpenAI's introduction of o3-pro signifies a significant step in AI reasoning, providing advanced functionalities for critical applications while highlighting ongoing improvements and market competition in AI development.