Nvidia has agreed to acquire Hugging Face, the platform where developers share and run open AI models, for $12,930,300,000. Chief executive Jensen Huang announced the deal in a company blog post on 3 September 2026.
According to reports of Nvidia's filing with the US Securities and Exchange Commission, about $11.9 billion goes to Hugging Face's shareholders and up to $1 billion is set aside as a stock retention programme for employees who join Nvidia. The deal requires regulatory approvals and other customary conditions, and Nvidia is aiming to complete it in the first half of 2027.
Hugging Face will keep operating under its own brand. Huang said its co-founders Clem Delangue, Julien Chaumond and Thomas Wolf approached him about the company's next chapter, and Wolf told Bloomberg Television that the three will stay on under a six-year retention arrangement.
What Hugging Face built
Delangue, Chaumond and Wolf founded Hugging Face in 2016. It started as a chatbot app for teenagers and changed course in 2017, when the founders began raising money, according to Bloomberg Línea. The company is headquartered in Brooklyn, New York, while most of its employees are based in France.
The platform has become the main public hub for open AI models. According to Nvidia, more than 18 million developers, researchers and creators use it to share more than 3 million models, 500,000 datasets and 1 million applications, and more than 200,000 companies use it to find, evaluate, customise and deploy AI. Nvidia says it is itself the largest contributor of open models and data to the site, with more than 500 models and 250 datasets published there.
“Hugging Face will remain an open platform for the entire AI ecosystem.”
Why Nvidia is buying
Huang wrote that AI moves faster when people can build together, and said Nvidia would help Hugging Face with reliability, safety, model evaluation, inference and deployment.
The central promise was neutrality. Huang said Hugging Face will remain an open platform for the whole AI ecosystem, that developers will keep choosing their own models, frameworks, clouds, inference providers and chips, and that Nvidia compute will not be required to build on or deploy through the platform. Support for open-source and open-weight models from all developers, and for multi-cloud and multi-accelerator work, will continue.
According to Le Monde Informatique, the purchase is the second-largest in Nvidia's history by announced value, after its $20 billion Groq deal in December. The announcement came about two weeks after reports that Hugging Face was exploring a sale at a valuation above $13 billion, slightly more than the final price.
The founders and the funding
Hugging Face's last funding round was a $235 million Series D in August 2023, led by Salesforce Ventures at a $4.5 billion post-money valuation. Google, Amazon, Nvidia, Intel, AMD, Qualcomm, IBM and Sound Ventures also took part. That round brought the company's total funding to about $395 million.
Other past investors include Sequoia Capital, Coatue and Lux Capital, which Le Monde Informatique described as one of the main beneficiaries of the sale. The outlet estimated that investors who put in less than $400 million in total could share around $7.3 billion in gains.
The Bloomberg Billionaires Index estimated that Delangue, 38, Chaumond, 42, and Wolf, 41, will each be worth about $1.8 billion after the deal. Delangue is chief executive, Chaumond chief technology officer and Wolf chief science officer.
What happens to users
The deal is not expected to close before 2027. Nvidia said the platform will keep its current operating model, including free uploading and downloading of models and datasets, and that the team will scale its work under the existing brand once the company is part of Nvidia.
Hugging Face did not disclose its headcount or revenue in the announcement. Le Monde Informatique noted that the deal still has to clear regulators, and Nvidia did not say which jurisdictions must approve it.
Chaumond, who previously co-founded the reading app Glose, later sold to Medium, pointed in a personal post on the deal to Huang's position on open-source AI and the way he defended it when it was under threat.
