OpenAI has launched O3-pro, a next-generation AI model it claims is its most powerful yet. A step up from its earlier “reasoning model” called O3, this new version is now available to ChatGPT Pro and Team users and promises improved performance in science, education, programming, business, and writing. While O3-pro introduces several advanced capabilities, it also comes with a few limitations — like slower response times and lack of image generation — as OpenAI continues to iron out technical issues.

Key Points:

• O3-pro is a reasoning model designed to work step by step through problems, improving reliability in math, physics, and code.
• It replaces o1-pro in ChatGPT Pro and Team starting Tuesday, with Enterprise and Edu users getting it next week.
• O3-pro is now live in OpenAI’s API, priced at $20 per million input tokens and $80 per million output tokens.
• The model can search the web, analyze files, reason visually, write Python code, and personalize answers via memory.
• In internal evaluations, O3-pro outperformed the previous O3 in all categories, especially for clarity and instruction-following.
• O3-pro scored higher than Google’s Gemini 2.5 Pro and Claude 4 Opus by Anthropic in benchmark tests like AIME 2024 and GPQA Diamond.
• Limitations include slower response times, disabled temporary chats, no image generation, and no Canvas feature support (yet).

Key Quotes:

“In expert evaluations, reviewers consistently prefer o3-pro over o3 in every tested category and especially in key domains like science, education, programming, business, and writing help.” – OpenAI changelog

“Reviewers also rated o3-pro consistently higher for clarity, comprehensiveness, instruction-following, and accuracy.” – OpenAI

Implications:

The release of O3-pro signals OpenAI’s focus on reasoning models — a shift aimed at reliability and depth rather than just speed or novelty. Its outperformance of major rivals like Gemini 2.5 Pro and Claude 4 Opus positions OpenAI as a continued leader in the AI arms race. However, the model’s current limitations and slower responses could impact usability for time-sensitive tasks. Still, with advanced tools and strong internal benchmarks, O3-pro could become the preferred model for developers, educators, and knowledge workers seeking precision over flash.

Source: https://techcrunch.com/2025/06/10/openai-releases-o3-pro-a-souped-up-version-of-its-o3-ai-reasoning-model/

AISQSquirrly

I am yourGateway

Gateway to AI Adoption. This is the Future of Marketing.