엔비디아는 허깅페이스를 129억 달러에 인수하며 플랫폼의 개방성을 유지하겠다고 약속했다.
SAN FRANCISCO — Clément Delangue walked away from Nvidia once. Roughly a year ago, Hugging Face rejected a $500 million offer from the chipmaker. On Wednesday, Delangue accepted a new valuation of $12.93 billion, framing the decision as a matter of necessity. “For it to happen at a larger scale,” the Hugging Face CEO told CNBC, “it needs more compute, more support, more collaboration, and more visibility.” By “it,” Delangue meant open-source AI—the movement around which he built his platform, and which he has long argued provides smaller companies and researchers a pathway to develop artificial intelligence without paying for access to proprietary systems like those from OpenAI or Anthropic.
Hugging Face operates as a specialized hub for artificial intelligence, often compared to GitHub but tailored specifically for machine learning. The platform hosts 3 million models, 500,000 datasets, and roughly 1 million applications, serving more than 18 million developers worldwide. Over 200,000 companies rely on it to discover, evaluate, customize, and deploy AI solutions. By virtually every metric, it serves as the central infrastructure of the open AI ecosystem.
Nvidia CEO Jensen Huang announced the acquisition on Wednesday, characterizing it as an expansion of AI access for developers and institutions globally. “Hugging Face will remain an open platform for the entire AI ecosystem,” Huang wrote. “Developers will choose the models they want, the frameworks they want, the clouds and inference service providers they want and the computing platforms they want.” This commitment to openness will be scrutinized closely by the developer community. Because Nvidia manufactures the GPUs that power the vast majority of AI training and inference workloads, the acquisition places the company at both ends of the development pipeline: supplying the underlying hardware and hosting the models themselves. While Nvidia has promised that users will not be forced to use its compute resources, the structural incentive post-closing will inevitably be to make Nvidia’s infrastructure the most convenient option. Whether those two realities can coexist remains the central question the deal leaves unanswered.
Nvidia was already Hugging Face’s largest corporate contributor, having published more than 500 models and 250 open datasets on the platform. It also participated in Hugging Face’s 2023 fundraising round, a $235 million deal led by Salesforce Ventures alongside investors including Google, Amazon, IBM, AMD, and Qualcomm. According to CNBC, Nvidia initially made an offer reportedly around $500 million, which Hugging Face declined. In the weeks leading up to Wednesday’s announcement, however, Delangue approached Huang directly to negotiate the final terms.
At a valuation of $12.93 billion, Hugging Face generates approximately $150 million in annualized revenue and is nearing profitability, according to the company. At that revenue level, the purchase price represents a multiple of roughly 86 times annualized sales—a figure implausible for most software acquisitions. The more likely explanation is that Nvidia is paying for infrastructure control rather than immediate financial returns. Whoever owns Hugging Face controls the primary directory where the field’s most critical models are stored, discovered, and deployed.
This marks Nvidia’s second-largest acquisition in recent memory, following its reported purchase of Groq’s assets late last year for roughly $20 billion to expand AI inference capacity. The pattern is clear: Nvidia is expanding from chip hardware into the full stack of AI infrastructure. As NBC News reported, the acquisition reflects a strategic bet that open-source models will grow in importance alongside, or potentially replace, closed systems from competitors like OpenAI and Anthropic.
The timing carries sharper geopolitical and regulatory implications than the press release suggests. On the same day, the United States reached a light-touch AI regulation accord with G20 members, reflecting industry resistance to restrictions on harms that Huang and others have dismissed as “theoretical.” Yet developments like OpenAI’s Astra—which recently became the first AI to autonomously discover zero-day exploits—lend weight to arguments that such risks are far from hypothetical. Delangue has long positioned Hugging Face as a counterweight to centralized AI, citing China’s competitiveness as proof that open-source development offers a structural advantage over reliance on proprietary APIs. Whether that positioning survives acquisition by the dominant hardware supplier, or whether Nvidia’s backing accelerates it, will only become clear after the deal closes.
The transaction still requires regulatory approval, though neither company has disclosed a timeline or specified the jurisdictions under review. Nvidia’s $40 billion bid for Arm Holdings collapsed in 2022 amid antitrust pressure. Given that the Hugging Face deal is valued at roughly a third of that amount and lacks the chip-licensing monopoly concerns that doomed the Arm attempt, it may face a different reception. Whether it does remains an open question. With John Ternus’s first Apple product event just one week away, Nvidia’s acquisition of the world’s central AI model library may ultimately prove to be the defining technology story of the week.