OpenAI refutes report that it will release AI model Orion later this year

OpenAI said on Friday that it does not plan to release an artificial intelligence model code-named Orion this year, refuting recent reporting on the company’s product roadmap.

The Verge reported on Thursday that the Microsoft (NASDAQ:MSFT)-backed startup plans to launch its next big AI mode in December, with the OpenAI granting access to its trusted partners first before a wider release through ChatGPT.

However, OpenAI spokesperson Niko Felix told The Verge that the company doesn’t “have plans to release a model code-named Orion this year” but that “we do plan to release a lot of other great technology.”

OpenAI chief executive Sam Altman called the report “fake news out of control” in a post on X.

According to the report, Microsoft (MSFT) engineers are working to host Orion on Azure as early as November.

While OpenAI views the new model as a successor to GPT-4 internally, it is not yet clear if the company will label it GPT-5 externally, the report added.

Earlier in the month, OpenAI became the second-highest valued startup in the world, after completing its latest fundraising round of $6.6 billion, propelling the company to a valuation of $157 billion.

The AI startup unveiled its latest flagship model, GPT-4o, in May, with improved capabilities across voice, text and imaging. It released a smaller, more cost-effective version, GPT-4o mini, in July.