Nvidia Buys Hugging Face for $12.9 Billion, Vows to Keep the Platform Open
News Summary
Nvidia has signed a definitive agreement to acquire Hugging Face, the widely used open-source AI model and dataset hub, in a deal valued at approximately $12.93 billion. The companies confirmed the transaction on September 3, 2026, Eastern Time, following days of media reports about the negotiations. It is Nvidia's second-largest acquisition on record, trailing only its roughly $20 billion purchase of chip startup Groq's assets in December 2025, and it marks one of the most significant consolidations yet in the open-source AI software ecosystem.
The Deal at a Glance
According to the definitive agreement signed on Wednesday, September 2, 2026, Nvidia will pay approximately $11.9 billion in cash and stock to Hugging Face shareholders, with an additional pool of up to $1 billion set aside in equity-based retention awards for Hugging Face employees, bringing the total transaction value to roughly $12.93 billion. Hugging Face's platform is a central hub for the AI development community, hosting more than 3 million AI models, over 1 million applications, and roughly 500,000 datasets, used by more than 18 million developers worldwide.
The valuation stands out because Hugging Face's annualized revenue was reported at only around $150 million as of last month, meaning Nvidia is paying a steep premium relative to current sales. Hugging Face had previously raised more than $395 million in venture funding, including a $235 million round in 2023 led by Salesforce Ventures. Notably, Hugging Face turned down a much smaller offer from Nvidia, reported at around $500 million, roughly a year earlier, underscoring how quickly the company's perceived strategic value has risen.
How the Deal Came Together
Speaking to CNBC on Thursday, September 3, 2026, Hugging Face co-founder and CEO Clément Delangue said his company approached Nvidia founder and CEO Jensen Huang about a deal over the summer of 2026, describing it as a moment when open-source AI development was "at a turning point" and needed more resources, scale, and visibility to keep pace with well-funded proprietary AI labs. Delangue said talks moved quickly once the two sides began discussing terms, telling CNBC that Nvidia felt like "a perfect home" for Hugging Face's mission.
In a joint statement, Huang said Hugging Face would continue to operate as an open platform after the deal closes: "Hugging Face will remain an open platform. Developers will choose the models, frameworks, clouds and computing platforms they want." Delangue echoed that framing, saying the backing of Nvidia would give Hugging Face "more compute, more support, more collaboration, and more visibility" as it scales to serve a growing global developer base.
What Happens to the Hugging Face Platform
Both companies have emphasized that Hugging Face will keep functioning as a neutral, multi-vendor hub rather than being folded into a closed Nvidia product line. Existing commitments to interoperability are expected to continue, and Nvidia has already contributed more than 500 models and roughly 250 datasets to the Hugging Face Hub as an outside partner, a relationship the acquisition is expected to deepen rather than replace. For the platform's global user base of researchers, students, and independent developers, the immediate expectation is continuity of service alongside expanded computing resources drawn from Nvidia's infrastructure.
Analysts covering the deal note that Nvidia's rationale extends beyond software: owning the leading distribution point for AI models gives the company a direct channel to sell computing capacity to the developers building and deploying those models, reinforcing the full-stack strategy Nvidia has pursued across hardware, software, and now open-source model distribution.
Strategic Context: A Broader Acquisition Pattern
The Hugging Face deal fits into a pattern of large, fast-moving investments Nvidia has made across the AI stack over the past year. The company has committed roughly $50 billion in aggregate investments and partnerships with frontier AI research labs, and in recent months struck a $6 billion collaboration with coding-focused AI startup Poolside to support open model development. The $20 billion Groq transaction in December 2025, structured primarily as an asset purchase and talent transfer rather than a full corporate acquisition, remains Nvidia's largest deal to date, with Groq's low-latency inference chips reported to be coming online in Nvidia's data center architecture later this year.
Regulatory Path Ahead
The transaction is subject to customary regulatory approvals before it can close, including a mandatory antitrust filing under the Hart-Scott-Rodino Act in the United States and an anticipated merger review by European Union competition authorities. Nvidia and Hugging Face have indicated they expect the deal to close in the first half of 2027, pending those reviews. As of the September 3, 2026 announcement, no regulator had issued a public statement on the transaction.
Because Nvidia already supplies the large majority of high-end AI training and inference hardware, industry observers expect reviewers to focus on whether combining that hardware position with ownership of the leading open model-distribution platform could disadvantage developers or hardware competitors relying on Hugging Face's neutrality, including maintained integrations with alternative chip platforms. Both companies have pointed to Huang's public commitment to keep the Hub open to all frameworks, clouds, and hardware providers as a response to that concern, though formal regulatory review will ultimately determine how the deal proceeds.
What It Means for Developers
For the millions of developers, students, and researchers who rely on Hugging Face daily to find and share models and datasets, the companies' public message has centered on stability: no near-term changes to how the platform is used, paired with promises of expanded compute access as Nvidia integrates its infrastructure behind the scenes. Educators and open-source contributors will likely watch closely over the coming months to see whether that openness commitment holds as the deal moves through regulatory review toward its expected close in 2027.