Openai effectively canceled the release of O3Which was scheduled to be the company’s next main AI model, in favor of what the Director General Sam Altman calls a “simplified” product offer.
In a a Post X on WednesdayAltman said that in the coming months, Openai will publish a model called GPT-5, which “integrates many [OpenAI’s] Technology, ”including O3, in its AI-operated chat Chatgpt and api. As a result of this roadmap decision, Openai is no longer planning to release O3 as a standalone model.
The company originally said in December that it was planning to launch an O3 sometime early this year.
“We want to do a better job to share our intended road page, and much better work simplifying our product offers,” Altman wrote in the post. “We want AI to ‘just work’ for you; we realize how complicated our model and product offers got. We hate the model selector [in ChatGPT] No matter how much you do and want to return to magical unified intelligence. ”
Altman has also announced that Openai plans to offer unlimited chat access to GPT-5 at the “standard smart configuration”, subject to “abusive thresholds”, after the model is generally available. (Altman refused to give more details about what this configuration-and these abusive thresholds-bring.) Subscribers to Chatgpt Plus will be able to operate GPT-5 at a “higher level of intelligence,” Altman said, while Chatgpt Pro subscribers will be able to run GPT-5 at “an even higher level of intelligence.”
“These models will incorporate voice, canvas, search, deep research and more,” Altman said, alluding to various functions that Openai has launched in Chatgpt over the past few months. “[A] Best goal for us is to unite [our] Models creating systems that can use all our tools, know when to think for a long time or not, and are generally useful for a very wide range of tasks. ”
Before GPT-5 launches, Openai in the next few weeks will release GPT-4.5, called “Orion”, which Altman says it will be the last “non-chain-thought-out model.” Unlike O3 and the other so-called O3 and Openai models, non-chain-of-thought models tend to be less reliable in domains such as mathematics and physics.
