Nvidia Agrees to Buy Hugging Face for $12.9 Billion

TL;DR

Nvidia, the company that makes most of the world’s AI chips, is buying Hugging Face, the platform millions of developers use to share and download open-source AI models, for about $12.9 billion. A regulatory filing breaks that into roughly $11.9 billion in cash for shareholders and up to $1 billion set aside to keep Hugging Face staff at Nvidia. Nvidia says the platform will stay open to any hardware or cloud, not just its own, but that promise is not a legal condition of the deal disclosed so far. The purchase is not final: it needs regulatory approval and is not expected to close until the first half of 2027.

What happened

Nvidia has agreed to acquire Hugging Face for $12,930,300,000, according to the company’s own announcement. TechCrunch independently reported the same $12.9 billion figure, tracing it to Nvidia’s announcement and its own review of Hugging Face’s funding history on Crunchbase.

Nvidia’s Form 8-K, filed with the Securities and Exchange Commission, breaks the total down: about $11.9 billion payable to Hugging Face stockholders, subject to adjustments, plus up to roughly $1.0 billion in equity-based retention pay for Hugging Face employees who join Nvidia. The same filing dates the definitive agreement to September 2, 2026, and puts the expected close in the first half of 2027, pending customary closing conditions, including regulatory approval. The filing’s risk disclosures also flag legislative or regulatory measures that could restrict or disadvantage open-source AI models as a material risk to the transaction.

Nvidia describes the platform it is buying in specific numbers: more than 18 million developers, researchers and creators, hosting over 3 million models, more than 500,000 datasets and over 1 million applications, used by more than 200,000 companies. TechCrunch reports the same order of magnitude and adds that Hugging Face was generating $150 million in annualized revenue as of August 2026, a figure TechCrunch attributes to a report by The Information.

Nvidia CEO Jensen Huang wrote that “Hugging Face will remain an open platform for the entire AI ecosystem” and that “NVIDIA compute will not be required to build on or deploy through Hugging Face”. Huang’s post also states that Nvidia is already the platform’s largest contributor of open models and data, having released more than 500 models and over 250 open datasets on Hugging Face.

Hugging Face was founded in 2016 and had raised over $395 million to date, per Crunchbase, with a 2023 round of $235 million led by Salesforce Ventures and joined by Google, Amazon, IBM and Nvidia. In a post on X, Hugging Face CEO Clem Delangue said the company approached Nvidia because scaling its open-source approach required more resources: “But for it to happen at [a] larger scale, it needs more compute, more support, more collaboration, and more visibility. That’s why we went to talk to Jensen, who offered to do exactly that with us.” TechCrunch also reports, citing the Financial Times, that Hugging Face turned down a $500 million acquisition offer from Nvidia last year.

Yahoo Finance reported that the deal is Nvidia’s second-largest, after a $20 billion non-exclusive agreement in December to buy assets from chipmaker Groq.

What this means (and what it does not)

The deal puts the dominant supplier of AI chips in control of the platform much of the open-source AI world uses to publish and download models. Nvidia’s own SEC filing names this concentration as a risk: it warns that laws or regulations restricting open-source AI could hurt the transaction, which is itself an acknowledgment that open-source AI’s regulatory standing now bears directly on Nvidia’s business. Hugging Face’s shareholders, who stand to receive roughly $11.9 billion for a company that had raised about $395 million from outside investors, have a direct financial stake in this being read as a continuation of Hugging Face’s mission rather than a loss of its independence. Google, Amazon and IBM, which put money into Hugging Face’s 2023 round and compete with Nvidia in AI infrastructure and cloud computing, now hold stakes in a company owned by that competitor.

What the announcement does not establish is that Hugging Face’s openness is locked in. Huang’s commitments — that developers can keep choosing their own models, frameworks, clouds and compute, and that Nvidia hardware won’t be required — are statements from the buyer, made before the deal has closed. Nothing disclosed so far describes a binding mechanism that would prevent those terms from changing once the acquisition is complete.

What we still do not know

Whether Nvidia’s openness commitment is contractually binding or simply a stated intention that could change after the deal closes. Whether hosting costs, compute subsidies or other economics for developers on Hugging Face will change under Nvidia ownership. Whether Nvidia will gain privileged access to model, dataset or usage data generated on the platform. What happens to Hugging Face’s paid Enterprise tier and its reported $150 million in annualized revenue. Which regulators will review the deal and on what timeline — the SEC filing cites required approvals as a closing condition without naming the agencies involved. Whether rival chipmakers and cloud providers change how they use or contribute to a platform now owned by a competitor. And what formal roles Hugging Face’s current leadership, including Clem Delangue, Julien Chaumond and Thomas Wolf, will hold inside Nvidia once the acquisition closes.