OpenAI o3-pro: The Latest Advanced AI Reasoning Model

OpenAI has officially introduced o3-pro, positioning it as the most sophisticated AI reasoning model to date. This cutting-edge o3-pro is being made available to ChatGPT Pro and Team users, superseding the previous o1-pro model. Access for Enterprise and educational users is anticipated to follow within one week.

Capabilities of OpenAI o3-pro

The newly launched model is part of the o3 family that OpenAI introduced earlier this year. It is specifically designed to excel in complex reasoning tasks across various fields, including science, education, mathematics, programming, and writing. This launch comes on the heels of an 80% price reduction for both o3 input and output costs, as indicated by CEO Sam Altman.

Access and Pricing

OpenAI o3-pro is readily available in the model picker for Pro and Team users starting today. Enterprise and educational users are expected to gain access the following week. As o3-pro operates on the same foundational model as o3, complete safety details can be found in the o3 system card.

Users can access o3-pro via both ChatGPT and its API, with a pricing structure set at $20 per million input tokens and $80 per million output tokens through the API.

Performance Insights

In internal evaluations, OpenAI reports that o3-pro has surpassed its non-pro counterpart across several critical domains. Human testers chose o3-pro for 66.7% of personal writing tasks and 62.7% of computer programming tasks, based on company data. Reviewers have also rated it more highly for clarity, instruction-following, and overall comprehensiveness.

Enhanced Functionality and Features

The o3-pro model incorporates tools that bolster its capabilities. These tools include web searches, file analysis, Python code execution, computer vision capabilities with reasoning, and access to user memory for delivering personalised responses. However, OpenAI notes that the use of these tools may result in responses taking slightly longer than those from o1-pro. The company recommends prioritising this model when accuracy is paramount, even if it means sacrificing some speed.

Limitations of o3-pro

Despite the enhancements, o3-pro does have certain limitations. The model cannot generate images, and temporary chats within ChatGPT are currently disabled due to a technical issue. Furthermore, the Canvas workspace feature is not supported by this model.

Future Developments and Competitive Positioning

OpenAI asserts that o3-pro has demonstrated strong performance in internal benchmark tests. However, comprehensive head-to-head benchmark data comparing o3-pro with leading competitors has yet to be released.

The introduction of o3-pro solidifies OpenAI’s position in the competitive landscape of high-performance AI models, as the company continuously refines its offerings for both general and professional applications.