Nvidia confirms $12.93 billion acquisition of Hugging Face
Nvidia is acquiring Hugging Face for $12.93 billion, pairing its AI computing reach with the leading open-model platform used by millions of developers.
Nvidia has confirmed that it will acquire Hugging Face for $12.93 billion, turning months of reported negotiations into one of the largest acquisitions in the AI infrastructure market. The deal would put the company behind a major open-model community directly under Nvidia’s ownership as developers and enterprises increasingly look beyond closed AI APIs.
Hugging Face says its platform hosts more than three million models, one million applications, and half a million datasets used by more than 18 million developers. Those figures were reported by TechCrunch from Nvidia and Hugging Face statements and should be treated as company-reported adoption metrics. Yahoo Finance, The Wall Street Journal, and the Financial Times separately reported the transaction at approximately $13 billion, but their available reports did not provide additional deal terms.
Nvidia is buying a critical layer of the AI stack
Founded in 2016, Hugging Face has become a central model hub for sharing, discovering, evaluating, and deploying machine-learning systems. Its repositories span open-weight language models, computer-vision systems, datasets, and developer applications. For builders, the platform often serves as the connection between research releases and production experimentation.
Nvidia already controls much of the hardware used to train and run advanced AI systems. Acquiring Hugging Face would extend that position into a software and distribution layer where developers choose models, tools, and deployment paths. It could also give Nvidia a direct channel for packaging compute capacity and AI services around the models that developers are already using.
TechCrunch reported that Hugging Face rejected a $500 million Nvidia offer last year. The platform’s latest financing round, in 2023, raised $235 million from investors including Salesforce Ventures, Google, Amazon, IBM, and Nvidia, according to the report. Crunchbase data cited by TechCrunch puts Hugging Face’s total funding above $395 million.
The acquisition price is significantly above those previous financing figures, reflecting the strategic value Nvidia appears to place on developer access and model distribution rather than only on Hugging Face’s reported revenue. The Information, as cited by TechCrunch, reported that Hugging Face was generating $150 million in annualized revenue last month. TechCrunch also reported that CEO Clem Delangue said in July that the company was approaching profitability, but the acquisition announcement did not disclose financial details.
Nvidia promises an open platform after the deal
Nvidia CEO Jensen Huang said Hugging Face would remain an open platform and continue supporting open-source and open-weight models. According to the statement reported by TechCrunch, developers would still be able to choose their models, frameworks, cloud providers, inference services, and computing platforms. Huang specifically said Nvidia hardware would not be mandatory for building or deploying through the platform.
That commitment will be closely examined because Nvidia’s commercial interests are clear. The company sells the GPUs, networking equipment, and software infrastructure that power much of the AI market. Ownership of Hugging Face could help Nvidia understand which models are gaining traction, influence preferred deployment patterns, and make its own infrastructure easier to adopt without formally restricting alternatives.
The distinction between an open platform and a neutral platform will matter to customers. Hugging Face can remain technically open while Nvidia promotes optimized paths for its chips, cloud partners, or inference services. For enterprise buyers, the practical questions will be whether model portability, access to competing hardware, and independent governance remain meaningful after the acquisition.
Huang has positioned open models as strategically important for the United States and the wider AI economy. TechCrunch reported that Nvidia has released more than 500 models and 250 open datasets on Hugging Face. Those are Nvidia’s own figures, and they indicate the company has already been using the platform as a distribution channel before taking ownership.
Evidence, claims, and what remains unknown
The acquisition itself is supported by Nvidia’s confirmation, as reported by TechCrunch, and by matching reports from Yahoo Finance, the WSJ, and the Financial Times. However, the available source material does not establish whether the transaction has closed, identify the exact structure of the consideration, or describe regulatory and shareholder conditions.
The largest platform figures—including more than three million models, one million applications, half a million datasets, and 18 million developers—are vendor-reported. They show the scale Hugging Face claims, but they do not reveal how many models are actively maintained, how many applications are used in production, or how much enterprise revenue comes from the platform.
The market context is nevertheless significant. Nvidia has been expanding beyond selling accelerators into model development and AI services. TechCrunch reported that the company recently struck a $6 billion agreement with coding startup Poolside to develop open models and said Nvidia had invested more than $50 billion in AI frontier labs. These figures and arrangements suggest the Hugging Face transaction is part of a broader strategy to shape the software ecosystem around Nvidia’s hardware position.
The deal also follows a period of tension between open and proprietary AI. Delangue described Hugging Face as an alternative to closed-source APIs, while TechCrunch reported that Nvidia’s open model helped the platform respond to cyberattacks after proprietary models failed to protect it. Those claims are based on executive comments and are not independently verified in the supplied evidence.
What the acquisition means for builders and enterprises
For AI developers, the immediate benefit could be deeper access to compute, evaluation tools, model hosting, and deployment services in one ecosystem. Nvidia could use its capacity more efficiently by connecting unused infrastructure with Hugging Face’s developer and enterprise offerings. TechCrunch identified that possibility as one commercial rationale for the deal.
For startups, the acquisition may reduce friction in moving from an open model experiment to a production service. It could also create more attractive optimization tools for Nvidia hardware. But builders that value cloud and hardware neutrality will need to test whether Hugging Face’s APIs, model formats, and deployment options remain equally usable across competing infrastructure.
Enterprise buyers face a similar trade-off. Nvidia ownership could bring stronger support, better performance tuning, and a clearer procurement path. It could also concentrate more of the AI supply chain in one vendor, increasing concerns about pricing power, lock-in, model governance, and access to independent alternatives.
The competitive impact will extend beyond Hugging Face. Cloud providers, model vendors, and enterprise AI platforms may respond by strengthening their own model registries, open-source programs, or multi-cloud deployment tools. Nvidia’s challenge will be to capture more value from the platform without weakening the neutrality that made it useful to developers in the first place.
What to watch next
The first signal will be Nvidia and Hugging Face’s formal explanation of the transaction: whether the deal has closed, how it will be financed, and whether regulators are expected to review it. Customers should also watch for changes to Hugging Face’s governance, pricing, model-hosting policies, and support for non-Nvidia hardware.
Product teams should track whether Nvidia introduces preferred inference services, GPU-optimized model releases, or bundled enterprise contracts. Researchers and open-source maintainers will want clarity on licensing, access to datasets, moderation policies, and whether Hugging Face continues to host models from Nvidia’s competitors on comparable terms.
Finally, reported revenue, profitability, developer activity, and production usage will matter more than headline repository counts. Those indicators will show whether Nvidia has acquired a durable software business or primarily a strategic distribution network for its AI infrastructure.
Creati.ai perspective
Nvidia’s purchase of Hugging Face is important because it combines control of AI computing with influence over the place where much of the open-model ecosystem is discovered and deployed. The transaction could accelerate the path from model release to production, but its value will depend on preserving choice rather than merely using openness as a route into Nvidia’s hardware stack.
For builders and enterprise buyers, the right response is not to assume that the platform has become closed or neutral. They should test portability, pricing, governance, and performance across multiple clouds and hardware providers as the integration develops. The acquisition’s long-term credibility will be measured by how much independence Hugging Face retains after becoming part of Nvidia.
