OpenAI has launched o3-pro, a new AI model designed for complex tasks demanding reliable performance. It is currently available to ChatGPT Pro users and via API integration, with Enterprise and Education users gaining access the following week. Built upon the o3 architecture, o3-pro surpasses both o3 and the earlier o1-pro in capabilities, according to OpenAI's benchmarks. Reviewers have consistently favoured o3-pro for its clarity, instruction-following, and overall accuracy, particularly in science, education, programming, business, and writing.
The model's key advantage lies in its enhanced reliability, demonstrated through rigorous evaluations requiring consistently correct answers across multiple attempts. This reliability is achieved through increased access to tools, including web search, Python, file analysis, and memory personalization, which results in longer response times. Some features, such as temporary chats and image generation, are temporarily unavailable. OpenAI is also cutting the price of its flagship o3 model by 80% to compete with other AI models in the market.
Related Articles
Altman envisions perfect, tiny AI
Read more about Altman envisions perfect, tiny AI →Manus Enters Text-to-Video Arena
Read more about Manus Enters Text-to-Video Arena →AI Defies Shutdown Orders
Read more about AI Defies Shutdown Orders →AI Training Content Market Emerges
Read more about AI Training Content Market Emerges →