The release of OpenAI's open model will take longer, its CEO Sam Altman said, citing the need for additional safety testing. 

The launch of OpenAI's much awaited open model has already been postponed once earlier.

In an X post, Altman last week said that the launch was expected sometime this week.

"We are delaying it; we need time to run additional safety tests and review high-risk areas," he said.

He added that the company is not yet sure how long it will take to get the model out.

In another X post, Aidan Clark, VP of research, OpenAI, who is leading the open model team said, "Capability wise, we think the model is phenomenal -- but our bar for an open source model is high and we think we need some more time to make sure we’re releasing a model we’re proud of along every axis."

Altman has been building the excitement around the company's open source model. In June he had remarked that the company had achieved something “unexpected and quite amazing,” without elaborating what that was.

The open source model space has become competitive with a string of AI startups in the US and China bracing to launch their own models.

Just last week Chinese AI startup Moonshot AI launched Kimi K2. Last month, Mistral launched its first family of AI reasoning models, called Magistral. Earlier in April, Chinese AI lab Qwen released a family of hybrid AI reasoning models. 

Apart from the release of OpenAI's open source model, there is already a big hype around the release of its GPT-5.

Last week, Elon Musk's AI startup xAI released its newest version of chat bot Grok, called Grok 4. Grok was launched by Musk in November 2023 as an answer to ChatGPT and Google's Gemini.