Nvidia agrees to buy Hugging Face for about $12.9 billion
The chip giant is acquiring the home of open-source AI models and datasets. The deal, its second-largest ever, is expected to close in the first half of 2027.
Nvidia has agreed to acquire Hugging Face, the platform where much of the world’s open-source AI work is shared, for about $12.9 billion.
The deal
- Price: about $12.93 billion, according to reports.
- Timing: expected to close in the first half of 2027, subject to regulatory approval.
- Scale: it would be Nvidia’s second-largest purchase, after the roughly $20 billion it paid for Groq assets at the end of last year.
Hugging Face’s chief executive told CNBC that the company approached Nvidia’s Jensen Huang weeks before the deal was agreed.
What Hugging Face is
If GitHub is where programmers share code, Hugging Face is where AI developers share models, datasets and demo applications. Hundreds of thousands of open models, from tiny research experiments to leading open-weight releases from major labs, are hosted there, along with the tools many developers use to download, run and fine-tune them.
It has become core infrastructure for anyone building with open AI models, including startups, universities and large companies.
Why Nvidia wants it
Nvidia sells the chips most AI runs on. Owning the place where developers find and run models brings it closer to the software layer, and to the millions of developers who choose which hardware their projects target. Nvidia has spent years building software, from CUDA to its inference tools, to make its chips the default choice. Hugging Face would extend that reach.
The questions
Hugging Face’s value comes partly from being neutral: a place where models built for any hardware, and by any company, sit side by side. Developers will be watching for any sign that the platform starts favouring Nvidia’s chips or software, and regulators are likely to ask the same question before approving the deal.
