On Tuesday, OpenAI unveiled its latest artificial intelligence (AI) model, named o3-pro. This model is touted as the company’s most sophisticated reasoning AI and is exclusively accessible to premium tier subscribers. Users can access o3-pro through ChatGPT and the company’s application programming interface (API). The new model reportedly offers enhanced performance over the previously released o3, particularly in fields such as science, education, programming, data analysis, and writing. This launch follows an announcement by OpenAI CEO Sam Altman, who revealed an 80 percent discount on input and output prices for the o3 model.
OpenAI Releases o3-pro For ChatGPT Pro Subscribers
The announcement of the o3-pro model was shared on X (formerly Twitter) via OpenAI’s official account. At present, it is only available to ChatGPT Pro subscribers and developers through the API, with enterprise and educational users set to gain access one week later.
For Pro subscribers, the o3-pro model will take the place of the o1-pro model in the selection menu. This launch comes nearly two months after OpenAI introduced the entry-level o3 model along with the o4-mini variant.
According to OpenAI, the new o3-pro model outperforms its predecessor and other non-pro versions in several domains. Internal evaluations indicated that the o3-pro had a 66.7 percent win rate in personal writing tasks compared to the o3 model, and achieved a 62.7 percent win rate in computer programming against the baseline model. However, the firm did not provide any benchmark comparisons with rival models like Gemini 2.5 Pro or Claude Opus 4.
The o3-pro model also includes advanced capabilities, such as the ability to use various tools including web search, file analysis, computer vision integrated with reasoning, Python programming, and accessing memory to provide personalized responses. However, OpenAI noted that the introduction of these tools results in slightly slower response times compared to the o1-pro model. Users are therefore encouraged to opt for the o3-pro model when the accuracy of responses takes precedence over speed.
There are certain limitations associated with the o3-pro model. OpenAI indicated that it cannot generate images and currently does not support temporary chats or Canvas functionality. The temporary chat feature has been disabled due to a technical issue, which the company hopes to resolve soon.
