Sam Altman: GPT-5 Will Be Delayed to Get Better — But OpenAI’s ‘Thinking’ AI o3 Is Coming Soon

OpenAI had planned a wide-scale launch of its o3 reasoning model in February, but then delayed it indefinitely. Now, the company is adamant that o3 and its next-generation o4-mini will be publicly available within a “couple of weeks,” OpenAI CEO Sam Altman said on social media platform X. He also explained why GPT-5 will only be available “in a few months.”

Image source: unsplash.com

Altman says the shift is part of the company’s commitment to “making GPT-5 much better than we originally thought.” GPT-5 is intended to be a unified model that includes what it calls reasoning capabilities. “We’ve also found that integrating everything seamlessly is more challenging than we thought. And we want to make sure we have enough capacity to support what we expect to be unprecedented demand,” Altman said.

Once the model is released, OpenAI plans to offer unlimited access to it at “default intelligence settings,” subject to “abuse thresholds.” ChatGPT Plus subscribers will be able to run GPT-5 at a “higher level of intelligence,” while ChatGPT Pro subscribers will be able to run GPT-5 at an “even higher level of intelligence,” Altman said.

«“GPT-5 will include voice, canvas, search, deep learning, and more,” Altman said earlier this year. “The big goal for us is to combine [our] models, creating systems that can use all of our tools, know when to think long and when not to think long, and generally be useful for a very wide range of tasks.”