Nvidia buys Hugging Face for $12.9B

The swirling rumors of the past few weeks just turned into one of the biggest acquisitions in AI history.

Nvidia has officially agreed to acquire open-source AI hub Hugging Face for a whopping $12.93 billion. CEO Jensen Huang announced the deal today, marking a massive consolidation of the open-source AI ecosystem under the hardware giant's roof.

To put that price tag in perspective, Hugging Face rejected a $500 million offer from Nvidia just last year, according to the Financial Times. The Information reported last month that the platform was generating around $150 million in annualized revenue, and CEO Clem Delangue recently mentioned they were getting close to profitability. Nvidia is paying an astronomical premium to seal the deal.

The open-source pledge

Hugging Face currently hosts over 18 million developers, 3 million models, 500,000 datasets, and 1 million applications, with more than 200,000 companies relying on it. The immediate panic from developers was whether the open platform would suddenly become an Nvidia-only walled garden.

Huang was quick to counter that narrative. He pledged that Hugging Face will remain an open platform, supporting multi-cloud, open-weight models, and alternative accelerators. In his words, Nvidia compute will not be required to build or deploy on Hugging Face.

Delangue echoed this on X, writing that while open source proved it could rival closed APIs, reaching the next level requires massive compute, support, and visibility. That realization led him directly to Jensen.

Still, Nvidia is already Hugging Face's largest open contributor, having dropped more than 500 models and 250 datasets on the site. TechCrunch noted that owning the ecosystem lets Nvidia tailor tools to its chips while packaging unused enterprise compute directly with Hugging Face's platform.

Why it matters

For everyday developers, this is not an immediate emergency. If Nvidia keeps its word, you will still be able to pull open-weight models, deploy on whatever cloud you want, and enjoy better infrastructure backing the platform.

However, the community's skepticism is completely justified. Hugging Face was the neutral town square of AI. Even if Nvidia does not actively block AMD or custom ASICs, it has every incentive to make the default experience smooth on CUDA and painful everywhere else. We think keeping Hugging Face genuinely platform-agnostic will be much harder in practice than Huang's blog post suggests.

For the wider industry, this cements Nvidia's shift from being a merchant of hardware to becoming the ultimate gatekeeper of AI development. Alongside its reported $6 billion partnership with Poolside and over $50 billion funneled into frontier labs, buying Hugging Face locks in millions of builders before they ever decide which hardware to buy. It is a brilliant, aggressive move to monopolize the developer workflow.