OpenAI Unveils o3-pro: A Powerful Upgrade to Its AI Reasoning Model
OpenAI has officially launched o3-pro, a significant upgrade to its AI reasoning model, o3. The company claims that o3-pro is its most capable model to date, designed to enhance performance in complex problem-solving tasks.
O3-pro builds upon the foundation of o3, which was released earlier this year as a reasoning model. Unlike conventional AI models, reasoning models tackle problems step by step, leading to more reliable outcomes in areas such as physics, mathematics, and coding. This approach makes o3-pro a valuable tool for professionals and researchers dealing with intricate challenges.
Starting Tuesday, ChatGPT Pro and Team users can access o3-pro, which replaces the o1-pro model. Enterprise and Edu users will gain access the following week, according to OpenAI. Additionally, o3-pro is now available in OpenAI’s developer API.
The pricing for o3-pro in the API is set at $20 per million input tokens and $80 per million output tokens. To put this in perspective, a million input tokens roughly equate to 750,000 words.
OpenAI emphasizes the superior performance of o3-pro over o3 in various categories. “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,” the company stated in a changelog. Reviewers also highlighted o3-pro’s enhanced clarity, comprehensiveness, instruction-following, and accuracy.
One of the key advantages of o3-pro is its access to tools that enable it to search the web, analyze files, reason about visual inputs, utilize Python, and personalize responses using memory. However, OpenAI notes that the model’s responses may take longer to generate compared to o1-pro.
Despite its advancements, o3-pro has some limitations. Temporary chats within ChatGPT are temporarily disabled while OpenAI addresses a technical issue. Additionally, the model cannot generate images, and it does not support Canvas, OpenAI’s AI-powered workspace feature.
On a positive note, o3-pro has demonstrated impressive results in AI benchmarks. In internal testing, it outperformed Google’s Gemini 2.5 Pro on AIME 2024, a math skills evaluation. It also surpassed Anthropic’s Claude 4 Opus on GPQA Diamond, a test assessing PhD-level science knowledge, showcasing its capabilities in advanced reasoning and knowledge application.
