OpenAI has introduced the rollout of its newest language mannequin, o3-pro, to all ChatGPT Professional and Staff customers. The brand new mannequin replaces the earlier o1-pro and is now obtainable by means of the mannequin picker in ChatGPT. Enterprise and Edu customers will obtain entry subsequent week.
OpenAI o3-pro is out there within the mannequin picker for Professional and Staff customers beginning immediately, changing OpenAI o1-pro.
Enterprise and Edu customers will get entry the week after.
As o3-pro makes use of the identical underlying mannequin as o3, full security particulars could be discovered within the o3 system card.…— OpenAI (@OpenAI) June 10, 2025
The corporate acknowledged, “OpenAI o3-pro is rolling out now to all Professional customers in ChatGPT and within the API.”
In accordance with OpenAI, o3-pro builds on the identical underlying mannequin as o3 however reveals stronger efficiency throughout crucial domains.
o3-pro is rolling out now for all chatgpt professional customers and within the api.
it’s actually sensible! i didnt imagine the win charges relative to o3 the primary time i noticed them.— Sam Altman (@sama) June 10, 2025
“In knowledgeable evaluations, reviewers persistently want OpenAI o3-pro over o3, highlighting its improved efficiency in key domains—together with science, schooling, programming, information evaluation, and writing,” the corporate mentioned.
The mannequin additionally handed OpenAI’s “4/4 reliability” benchmark. The rigorous ‘4/4 reliability’ analysis considers a mannequin profitable provided that it solutions a query appropriately in all 4 makes an attempt reasonably than only one.
Like its predecessor, o1-pro, the brand new mannequin maintains robust tutorial capabilities. “Like OpenAI o1-pro, OpenAI o3-pro excels at math, science, and coding as proven in tutorial evaluations.”
OpenAI additional highlighted the mannequin’s software integrations. “OpenAI o3-pro has entry to instruments that make ChatGPT helpful—it may well search the net, analyse recordsdata, cause about visible inputs, use Python, personalise responses utilizing reminiscence, and extra.”
In the meantime, OpenAI has decreased the value of the o3 mannequin by 80%, with enter now priced at $2 per million tokens and output at $8 per million tokens. Alternatively, the o3-pro mannequin is priced at $20 per million tokens for enter and $80 per million tokens for output—87% cheaper than o1-pro.
OpenAI CEO Sam Altman introduced that the corporate is pushing again the discharge of its open-weights mannequin, which was initially anticipated in June. The mannequin will now be launched later this summer time.
In a submit on X, Altman indicated that the workforce could be taking extra time with their open-weights mannequin, suggesting that it might be anticipated later in the summertime, however not in June.
He added that the delay is because of a brand new growth from the analysis workforce, which he described as “surprising and fairly superb.” In accordance with Altman, “We predict it will likely be very, very definitely worth the wait, however wants a bit longer.”
The submit OpenAI Launches o3-pro, Delays Open-Weights Mannequin Launch appeared first on Analytics India Journal.