OpenAI launched its o3-pro synthetic intelligence (AI) mannequin on Tuesday. Dubbed as the corporate’s most superior reasoning mannequin, it is just out there to probably the most premium tier of subscribers. The o3-pro mannequin will likely be out there each in ChatGPT in addition to through the AI agency’s software programming interface (API). The mannequin is claimed to be extra performant in comparison with o3 in topic areas resembling science, training, programming, information evaluation, and writing. The AI mannequin was launched simply hours after OpenAI CEO Sam Altman introduced an 80 % low cost within the enter and output costs of o3.
OpenAI Releases o3-pro For ChatGPT Professional Subscribers
In a publish on X (previously often called Twitter), the official deal with of OpenAI introduced the rollout of the o3-pro mannequin. Presently it’s out there to solely ChatGPT Professional subscribers, and to builders through an API. Enterprise and Edu customers will get entry to it one week later.
The o3-pro AI mannequin will exchange the o1-pro mannequin within the mannequin picker choice for Professional subscribers. Notably, the flagship-tier reasoning mannequin from the o3 household arrives practically two months after the San Francisco-based AI agency launched the baseline o3 and the o4-mini fashions.
OpenAI says the newest reasoning mannequin outperforms the non-pro variant in a number of areas. Primarily based on inside evaluations with human testers, the corporate stated the o3-pro achieved a 66.7 % win price in private writing over the o3 mannequin. It additionally achieved 62.7 % win price towards the baseline mannequin in laptop programming. Notably, the corporate didn’t share any benchmark assessments evaluating the mannequin towards opponents resembling Gemini 2.5 Professional or Claude Opus 4.
The o3-pro AI mannequin can be able to instrument use, and it will possibly entry instruments resembling internet search, file evaluation, laptop imaginative and prescient with reasoning, use Python, entry reminiscence to personalise responses, and extra. Nonetheless, the corporate additionally highlighted that because of entry to instruments, the mannequin additionally takes barely longer than o1-pro to generate responses. As such, OpenAI is asking customers to decide on it when accuracy of a solution is extra necessary than pace.
There are some limitations with the massive language mannequin (LLM) as effectively. As per the OpenAI, the o3-pro mannequin can not generate pictures and doesn’t assist short-term chats or Canvas. Notably, short-term chats are disabled for the mannequin because of a technical subject, and could possibly be fastened quickly.
