Nvidia is reportedly buying one of AI’s biggest model-sharing platforms, potentially extending its reach beyond chips and deeper into the software ecosystem.
The company has agreed to acquire Hugging Face for $12.9 billion, The Information reported Wednesday, citing a person familiar with the deal. The companies have not publicly confirmed the transaction, and reports indicate the agreement could still fall apart.
Hugging Face runs a widely used platform where developers share, test, and build AI models and datasets. The company was founded in 2016 and has become a central hub for open-source AI, hosting millions of models and datasets.
The reported price is a huge jump from Hugging Face’s last disclosed valuation.
A $235 million funding round in 2023 valued the company at $4.5 billion. Nvidia participated in that round alongside investors including Salesforce, Google and IBM. The company also offered to invest $500 million in Hugging Face at a $7 billion valuation late last year, The Information reported, but Hugging Face reportedly rejected the offer because it did not want a dominant investor influencing its decisions.
Why this deal matters to Nvidia
The acquisition would extend Nvidia’s reach beyond GPUs into the software and model layer of AI.
That matters because OpenAI is developing custom processors, while Google and Amazon are expanding their own AI chip portfolios. Anthropic is also increasing its use of Amazon’s Trainium chips, giving major AI developers more alternatives to Nvidia hardware. A stronger open-source ecosystem could keep developers using infrastructure built around Nvidia hardware.
Hugging Face also sells paid model hosting, inference and enterprise services. Bringing the platform under Nvidia could give the chipmaker a more direct role in how developers deploy AI models and create more opportunities to direct workloads toward Nvidia infrastructure.
The economics are striking. Hugging Face recently reached about $150 million in annualized revenue, according to The Information. Based on that run rate, the reported purchase price would equal roughly 86 times revenue.
The neutrality problem
The deal could create an awkward challenge for Hugging Face. Its appeal comes partly from being a neutral platform supporting models and hardware from companies across the industry, including Nvidia rivals such as AMD and Intel. Nvidia ownership could make some developers question whether that neutrality will survive.
At the same time, the acquisition could give Hugging Face access to Nvidia’s enormous financial resources and help accelerate its push into open-source AI.
Hugging Face CEO Clément Delangue has publicly backed open models, particularly for AI cybersecurity. “AI cybersecurity is going to become a huge market in the U.S. and in the world,” he told CNBC in an interview published Aug. 27. “In this market, probably open models will be kings.”
More must-read AI coverage
A costly bet on the open AI economy
For Nvidia, the reported $12.9 billion price is less about Hugging Face’s current revenue than about gaining influence over a strategic gateway to AI developers.
The deal could be powerful but risky because Hugging Face’s value depends partly on openness and broad industry participation. Enterprise teams that rely on the platform should watch for changes to pricing, model availability, data handling and support for competing chips and cloud services. No such changes have been announced.
If completed, the acquisition would show Nvidia trying to secure its position through more than faster chips. The key question is whether it can expand Hugging Face without weakening the neutrality that made the platform valuable.
What this could mean for enterprise AI
For enterprise buyers, the bigger industry question is what happens as AI ecosystems become more tightly integrated. Vendor choices may increasingly influence which chips, clouds, development tools and deployment options are easiest to use together.
That shift could create real efficiencies. Tighter integration across the stack can simplify deployment, improve performance, and reduce some of the complexity of stitching together AI systems from multiple vendors. But it can also make switching providers harder if organizations become deeply tied to one company’s tools, interfaces and infrastructure.
The industry impact goes beyond Nvidia and Hugging Face. If more large AI vendors pursue similar strategies, enterprises could face a market where the biggest competitive battles are no longer just about who builds the best model or fastest chip, but who controls the broadest ecosystem around them.
For CIOs and IT leaders, the long-term issue is therefore less about one acquisition and more about preserving flexibility. As AI stacks become more integrated, organizations may need to pay closer attention to interoperability, portability and exit costs before committing heavily to a single vendor ecosystem.
Read more: OpenAI’s Jalapeño inference chip shows how AI developers are building custom processors to reduce costs and their dependence on Nvidia hardware.
Read the full article here
