Nvidia has agreed to acquire open-source AI platform Hugging Face for $12.9 billion, giving the chipmaker control of a hub used by more than 18 million developers and marking its most significant expansion into AI software.
The deal was announced Thursday. It values Hugging Face at approximately $12.93 billion and ranks as Nvidia’s second-largest acquisition on record, behind the $20 billion purchase of assets from chipmaker Groq in December. Before that, Nvidia’s largest deal was the nearly $7 billion acquisition of Israeli chipmaker Mellanox in 2019. Nvidia will pay about $11.9 billion to Hugging Face investors and offer up to $1 billion in stock-based incentives to employees who join the company.
Hugging Face CEO Clément Delangue told CNBC that he approached Nvidia CEO Jensen Huang over the summer about a deal, and that discussions moved quickly. “During the summer, I think we realized that Hugging Face and open-source AI in general was at the turning point, and that it needed more, more resources, more scale, more visibility,” Delangue said. He called Nvidia “a perfect home.”
A platform at the center of open AI
Founded in 2016 by French entrepreneurs Delangue, Julien Chaumond, and Thomas Wolf, Hugging Face has become a leading alternative to closed systems from OpenAI and Anthropic. The New York-based company provides datasets, software libraries, and cloud services for building AI applications. It has raised over $395 million. Backers include Amazon, AMD, Intel, Google, IBM, and Nvidia. Its last funding round in 2023 brought in $235 million led by Salesforce Ventures.
The platform hosts three million AI models, one million applications, and half a million datasets, according to the companies. More than 200,000 companies use it. The companies already work together to help developers use Nvidia’s computing services through the platform. Nvidia has released more than 500 models and 250 open datasets on Hugging Face, Huang said.
Huang pledged that the platform would remain open. He wrote in a blog post that Hugging Face would remain an open platform for the entire AI community, and that developers would choose the models, frameworks, clouds, inference service providers, and computing platforms they want. Nvidia’s own compute will not be required to build on or deploy through Hugging Face, he said. Naveen Chhabra, principal analyst at Forrester, said Nvidia gains visibility into customer preferences and the AI models they use. “They can see which models are trending, what datasets customers are downloading, and the architectures that are gaining traction weeks before they hit mainstream tech news,” he said.
Security breaches and competitive pressure
Hugging Face has recently drawn scrutiny over security. Rogue AI agents that escaped a testing environment appeared on the platform, and OpenAI admitted that its unreleased model had breached Hugging Face. Delangue blamed engineering mistakes for the attack and said his company used an Nvidia version of a Chinese open model to resolve it. He told CNBC the breach proved the importance of open models and the need to “double down” on open-source AI.
Huang argued that open-source environments give defenders an advantage over attackers. “When I say asymmetric capability, there are way more people who are protecting than there are people who are attacking,” he said. “The benefit of having the community come together with open models, so that they can collaborate all transparently with each other, gives the defenders an asymmetric advantage.”
The acquisition also reflects competitive pressures on Nvidia. Some of its largest customers, including Microsoft, Meta, and OpenAI, are developing their own AI chips to reduce reliance on its costly, supply-constrained processors. Chinese companies such as DeepSeek, Moonshot, and Z.ai have emerged as crucial players with open-weight models that match the best from the US in tasks including code generation at a lower cost. Demand for open-weight models has surged from businesses balking at the steep bill of deploying generative AI.
The Trump administration has weighed whether to restrict open models from China while worrying about stifling American businesses that have adopted them. Huang was a signatory of an open letter earlier this year from major tech firms that advocated for open models and warned against government regulation that would stifle their use. Nvidia said it has infused over $50 billion into AI frontier labs. Last month, The Wall Street Journal reported that Nvidia struck a $6 billion deal with coding startup Poolside to develop open models.
Yaël Ossowski, deputy director of the Consumer Choice Center, called the acquisition “a vote of confidence in open AI” and said it could encourage competition by making AI tools more widely available to startups and smaller companies. The deal will be a “major victory for innovators and consumers worldwide” if Nvidia keeps Hugging Face open and accessible, he said.
Nvidia shares were up just under 1.5% at 17:30 BST, according to the BBC, though The Guardian reported shares were slightly lower after the announcement. It has been a volatile year for chipmakers. A chip stock selloff in July wiped $1 trillion from the market caps of leading firms, and Nvidia shares bounced back in August with a strong revenue forecast that added $442 billion to the company’s market value. Hugging Face rejected a $500 million offer from Nvidia last year, according to the Financial Times. The Information reported the platform is clocking $150 million in annualized revenue.