Nvidia Reportedly Agrees to Buy Hugging Face for $12.9 Billion

Nvidia has reportedly agreed to buy Hugging Face, the open-source AI model repository often called the “GitHub of AI,” for $12.9 billion. The report comes from The Information, which cited a person familiar with the matter on Wednesday night. Neither Nvidia nor Hugging Face has confirmed the deal, and a competing report from Business Insider, published the same night, says the talks had not yet produced a signed agreement.

What’s Reportedly Being Paid, and What Comes With It

The $12.9 billion figure would be a striking jump for a company that, according to The Information, was generating about $150 million in annualized revenue as of earlier this week, up from roughly $100 million just two months before that. Hugging Face founder and CEO Clem Delangue told TechCrunch last month that the growth had brought the company “close to profitability.”

It would also be a steep markup on Hugging Face’s last known valuation. The company raised $235 million in a 2023 funding round that valued it at $4.5 billion, with investors including Salesforce, Alphabet’s GV, IBM, and Nvidia itself. Nvidia had already tried once before to get closer to Hugging Face: the Financial Times reported in January that Hugging Face turned down a $500 million investment from Nvidia last year that would have valued the company at $7 billion, Hugging Face said at the time it didn’t want an investor dominant enough to sway its decisions. An outright acquisition avoids that problem for Hugging Face’s leadership by removing the question of control altogether.

Two Reports, Two Different Pictures of How Final This Is

The two Wednesday-night reports don’t fully agree on how close the deal actually is. The Information described Nvidia and Hugging Face as having already agreed to terms. Business Insider, which first reported over the weekend that Hugging Face was fielding takeover interest, said the same night that talks valuing the company at more than $13 billion had not yet produced a signed agreement and could still fall apart. Tech Startups reported that Reuters also picked up the story, and that neither company responded to requests for comment. CNBC‘s write-through matched The Information’s $12.9 billion figure, while Bloomberg‘s report leaned on Business Insider’s more cautious account of the talks.

TechCrunch, summarizing both reports, noted that Nvidia’s silence here is itself notable: the company has moved quickly in the past to publicly correct reporting it considered inaccurate, and it has not done so in this case.

Why Nvidia Wants an AI Model Hub, Not Just Chips

The strategic logic, as described in The Information’s reporting, centers on defending Nvidia’s position at the top of the AI hardware stack. OpenAI, Google, Amazon, and Anthropic are all developing their own AI chips to reduce how dependent they are on Nvidia’s GPUs. A thriving open-source model ecosystem gives smaller developers and companies an alternative to those closed labs, and much of that ecosystem still runs on Nvidia hardware. Owning the place where that ecosystem gathers, rather than just supplying the chips underneath it, would give Nvidia more direct influence over how it grows.

There’s also a cloud-computing angle. Nvidia reportedly scaled back its own DGX Cloud service roughly a year ago, and The Information reported that owning Hugging Face, which already rents out computing power for developers to run their models, could give Nvidia a way back into that market without rebuilding it from scratch. Nvidia has also committed to help cover the cost of tens of billions of dollars in cloud-computing commitments made on its customers’ behalf; if some of that reserved capacity goes unused, Hugging Face’s customer base gives Nvidia somewhere to resell it.

Hugging Face’s Very Public Alignment With Nvidia

Delangue has spent much of this year positioning Hugging Face alongside Nvidia’s arguments for keeping open-weight AI models unrestricted, as Washington has debated whether to limit them on national-security grounds. Nvidia CEO Jensen Huang and Hugging Face were among the 25 companies that signed a joint letter this year urging the U.S. government not to impose broad restrictions on open-weight models, a push sxz.io covered in July. In a CNBC interview in late July, Delangue pointed to that same letter while warning that China is “clearly dominating” open-source AI, a debate fueled in part by Chinese labs like Moonshot AI, whose Kimi K3 model has matched leading U.S. systems on benchmarks at a fraction of the cost.

Delangue has also described Hugging Face’s own security posture in Nvidia’s terms. In an appearance on CBS’s “Face the Nation” this month, he said Hugging Face used an Nvidia-modified version of a Chinese open-source model to help defend itself after a cyberattack, an apparent reference to the intrusion sxz.io reported on in July, in which an unreleased OpenAI model breached Hugging Face’s systems during a security test.

What an Acquisition Would Change

Hugging Face’s value to developers has always rested partly on its neutrality: it hosts competing models and supports hardware from Nvidia’s rivals, including Google, Amazon, and Microsoft, all of which are building their own AI chips. Folding Hugging Face into Nvidia would put a company that many AI teams treat as neutral infrastructure under the ownership of the industry’s dominant chipmaker, and it’s not yet clear whether that changes how the platform treats models optimized for competing hardware.

For now, the deal remains unconfirmed by either party, and the two initial reports don’t even agree on whether an agreement has actually been signed. sxz.io will follow up if Nvidia or Hugging Face make an official statement.