OpenAI has delayed the release of its upcoming open-source AI model. The company’s CEO, Sam Altman, announced that more time is needed to complete safety tests. This is the second time the release has been pushed back, after an earlier delay in June.
Altman shared the update on X (formerly Twitter), saying, “We planned to launch our open-weight model next week. We are delaying it; we need time to run additional safety tests and review high-risk areas. We are not yet sure how long it will take us.”
This new model would be OpenAI’s first open-source release since 2019 and the first one focused on reasoning abilities. In April, OpenAI had asked developers to fill out a feedback form to learn how they plan to use the model.
The AI model was expected to be on the same level as OpenAI’s advanced o-series models. But unlike the company’s closed products like ChatGPT, this model would allow developers to download the weights, fine-tune it, and run it on their own systems.
However, OpenAI will not release the training methods or the design of the model. This means developers won’t be able to rebuild the model from scratch or create new models based on it.
Many other AI companies, such as Google, Meta, Anthropic, xAI, and Alibaba, have recently released open-source models. Alibaba’s Ernie models are the latest example. These releases help boost the open AI community and give developers more control and flexibility.
OpenAI’s planned open-source model was seen as a way to balance two goals: offering powerful tools to developers and keeping its most advanced models like GPT-5 for enterprise use. GPT-5 is also expected to launch soon and may include new features such as better reasoning and the ability to handle multiple types of inputs.
For now, though, OpenAI has not given a new date for when the open-source model will be available. Developers will have to wait until the company finishes its safety checks.
Also Read: Google Gemini App May Soon Get Search Bar on Android
