Nvidia now owns the platform that decides which AI models the world actually usesNvidia Buys Hugging Face
Published: September 4, 2026 | Updated: September 4, 2026 | Category: AI | By Mahesh
Editor's note: This confirms and updates the reporting Depth Grid covered on August 27, when The Information and Business Insider had published conflicting accounts of whether this deal had actually been finalized. It is now officially confirmed.
Nvidia CEO Jensen Huang confirmed on September 3 that the company has agreed to acquire Hugging Face for $12.93 billion, ending weeks of conflicting media reports about whether the deal was signed or still in negotiation. "I'm excited to announce that NVIDIA has agreed to acquire Hugging Face for $12,930,300,000," Huang wrote in a post on Nvidia's own corporate blog, the down-to-the-dollar figure itself a small signal of how precisely the transaction's terms were negotiated.[1] The confirmation came alongside an SEC filing disclosing the transaction's structure, according to Adweek's reporting, which reviewed the filing directly: roughly $11.9 billion payable to Hugging Face shareholders, plus up to $1 billion in retention equity earmarked for employees joining Nvidia as part of the deal.[2]
The deal is Nvidia's second-largest acquisition on record, trailing only the company's $20 billion purchase of assets from AI chip startup Groq in December, and ahead of its roughly $7 billion purchase of Israeli chipmaker Mellanox back in 2019, according to CNBC's reporting on the confirmed transaction.[3] The deal is expected to close in the first half of 2027, pending regulatory approval.
How the Deal Actually Came Together
The origin story, as both CEOs described it publicly this week, runs counter to the typical acquisition narrative of an acquirer pursuing a reluctant target. Hugging Face CEO Clément Delangue told CNBC directly that his company approached Huang, not the other way around. "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," Huang told CNBC's Becky Quick on Squawk Box, recounting the conversation. Delangue's own account, relayed in Benzinga's coverage of the announcement, was similarly direct: "We told him we want to make open-source AI big, and he told us, 'Let's do it.'"
This detail resolves a genuine open question from earlier reporting, and it is worth being explicit about what changed. Depth Grid's earlier coverage on August 27 noted that Hugging Face had previously turned down a $500 million investment from Nvidia in late 2025, valuing the company at $7 billion, specifically to preserve independence from any single investor. That rejection stood for months. What changed by summer 2026, according to Delangue's own account, was a specific incident: he cited the summer breach involving OpenAI agents as direct evidence that open-source AI needed to scale faster and with more resources than Hugging Face could marshal independently, arguing the company could not adequately defend itself while relying on proprietary, closed-source infrastructure alone.
Why Nvidia's Own Framing Is Doing a Lot of Work
Huang's public commitment to keeping Hugging Face neutral is the central claim skeptics of this deal will be watching most closely over the coming months, and reporting on the announcement has already surfaced real disagreement about whether that neutrality can actually hold once a single hardware vendor owns the platform outright. Tekonyx founder Sid Nag, quoted in Benzinga's coverage, said Nvidia's resources could genuinely accelerate open-model adoption, but warned Hugging Face could gradually become what he called an "Nvidia-centered distribution channel" regardless of the company's stated intentions. Linthicum Research founder David Linthicum went further, questioning whether the two companies' cultures, customer bases and business models actually fit together, predicting the promised "1+1=3" combination Nvidia is selling to the market could just as easily deliver something closer to "1+1=1.2."
There is recent precedent for exactly this kind of regulatory skepticism shaping how a deal like this gets scrutinized. Benzinga's reporting noted that Nvidia previously abandoned its proposed acquisition of chip design firm Arm after regulators argued that Nvidia owning neutral technology widely used by its own competitors could give it privileged access to sensitive competitive information and weaken rival innovation across the industry. Hugging Face occupies a structurally similar position in the AI software stack: it is the platform where Nvidia's own competitors, and the customers of Nvidia's own competitors, discover, test and deploy AI models, many of which have nothing to do with Nvidia hardware specifically. Whether antitrust regulators view this deal through that same Arm-shaped lens is likely to be the central question determining if the deal actually closes on the H1 2027 timeline Nvidia has targeted.
The Strategic Logic, According to Nvidia's Own Numbers
Nvidia's own blog post lays out the scale of what it is acquiring in specific figures: more than 18 million developers, researchers and creators use Hugging Face to share more than 3 million models, 500,000 datasets and 1 million applications, and more than 200,000 companies use the platform to discover, evaluate, customize and deploy AI. CFRA Research equity analyst Angelo Zino, quoted in Yahoo Finance's coverage, offered a framing worth taking seriously: "I think it's less about the financials, at least today or even over the next 12, 18, 24 months, [and] it's more along the lines of building that AI ecosystem and trying to develop more and more capabilities and technology out there for the AI ecosystem." In other words, this deal is not being priced primarily on Hugging Face's current revenue, which Depth Grid's earlier coverage noted sat around $150 million annualized, but on strategic control over AI model distribution itself.
That strategic logic connects directly to a defensive pressure Nvidia is facing from its own largest customers. Adweek's reporting on the confirmed deal noted explicitly that the acquisition lands as Nvidia's biggest customers work to reduce their reliance on its GPUs, with both Anthropic and OpenAI developing their own proprietary chips. Owning the distribution layer where open-source models get discovered and deployed gives Nvidia a form of influence over the AI ecosystem that persists even as its largest customers diversify away from its hardware specifically.
What This Means for the Broader Pattern Depth Grid Has Tracked
This confirmed acquisition slots directly into a series of Nvidia moves Depth Grid has covered closely over the past month. An SEC filing earlier this month revealed Nvidia holds a $21 billion equity stake in SpaceX and a $30 billion stake in Intel, both companies that separately pledged Nvidia chip exclusivity. Nvidia also announced a $500 billion Wall Street financing alliance to help fund AI infrastructure purchases by its own customers. Read together with this week's confirmed Hugging Face deal, a consistent strategy emerges: Nvidia is simultaneously taking equity positions in its largest customers, helping finance the infrastructure those customers buy, and now directly owning the software distribution layer that determines which AI models developers actually choose to run. Each move individually has its own stated rationale, but collectively they describe a company moving aggressively to control every layer of the AI stack it touches, not just the chips it has historically been known for.
Read More on Depth Grid
- Nvidia says it hasn't bought Hugging Face. Two outlets disagree about which one is right.
- Nvidia's $10 billion AI bet just turned into a $21 billion rocket company stake nobody knew about
- Nvidia just asked Wall Street for $500 billion. Six firms said yes in one meeting.
- OpenAI built an AI that can hack anything on its own. Then it had to decide whether to ship it.
- A company with zero operating data centers just filed to go public at a $50 billion valuation
Article by Depth Grid News Desk | depthgrid.in