After effectively cancelling the consumer launch of the O3 inference model in February, Openai says it is currently aiming to release both the O3 and its next-generation successor, the O4-Mini, in “a few weeks.”
In a post on X on Friday, Openai CEO Sam Altman said, of course, it is related to Openai’s upcoming GPT-5.
“(w)e could make the GPT-5 much better than we originally thought,” writes Altman. “(w)e felt it was more difficult than being able to integrate everything smoothly.
Altman added that he expects Openai to roll out GPT-5 “in a few months.”
As far as Openai has released details about GPT-5, the company says it intends to provide unlimited chat access to GPT-5 in “standard intelligence settings” that are subject to “abuse thresholds.” ChatGpt Plus customers can run GPT-5 with “higher level of intelligence,” while subscribers to Openai’s ChatGPT Pro plan can run GPT-5 with “higher level of intelligence.”
“(GPT-5) incorporates audio, canvas, search, deep search, and more,” Altman wrote in an X post earlier this year, referring to the various features Openai has launched in ChatGPT over the past few months. “(a) Our best goal is to create a system that can use all the tools, know if we should think about it for a long time, and integrate (our) models that are generally useful for a very wide range of tasks.”
Openai is facing increased pressure from rivals such as the Chinese AI Lab Deepseek, which took a “open” approach to launch the model. In contrast to Openai’s strategy, these “open” competitors make the model available to the AI community for experimentation and, in some cases, commercialization.
In addition to the O3, the “O3 Pro” models O4-MINI, and the GPT-5, OpenAI will debut its first open language model since GPT-2 in the coming months. The model has inference capabilities, Altman said earlier this week, and is subject to additional safety assessments.