Nvidia's $12.9 billion buys the model registry 200,000 companies deploy from

Jensen Huang announced in an NVIDIA blog post that NVIDIA has agreed to acquire Hugging Face for $12,930,300,000.

The roughly $12.9 billion total breaks down into approximately $11.9 billion in purchase price plus a stock program of up to $1 billion to retain staff.

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.

According to the SEC filing, the deal will not close until the first half of 2027 and still needs regulatory approval.

Delangue says Nvidia has committed to backing Hugging Face while keeping the platform open, independent and compute-agnostic, with the founders and existing team staying on.

Nvidia offers its own NVFP4 versions of Kimi K2.6 and GLM-5.1 tuned for its chips, and its own open models sit behind the strong Chinese models.

Huang says co-founder Clem Delangue came to him while considering the company's next chapter, while co-founder Thomas Wolf writes on LinkedIn that Huang made Delangue the offer to build the hub into an open, independent and hardware-neutral platform.

Nvidia says it is the largest contributor of open models and data to Hugging Face, with more than 500 models and more than 250 open datasets released there.

Microsoft bought GitHub for $7.5 billion in 2018 and similarly promised the platform would remain independent and open; GitHub largely retained that openness, though Microsoft's later use of public GitHub code to help train the proprietary, paid Copilot service sparked a backlash in parts of the open source community.

Hugging Face distributes models and also rents out compute time to run them; Nvidia scaled back its own DGX Cloud because it did not want to poach rental customers from its big buyers, and its Lepton marketplace now routes jobs to partners like CoreWeave, Lambda and Nebius, with Hugging Face plugging into it through a cluster service.

Google, Amazon and OpenAI are building their own accelerators, and more recently so is Anthropic, while open models run across many clouds, inside companies, at universities and in government agencies, a customer base that does not build its own chips.

Huang summarised the logic for Axios in a single line: "Free AI should be great for hardware."

Delangue argues Hugging Face has reached the point where keeping open source AI competitive requires substantially more resources: "It needs more compute, more support, more collaboration and more visibility."

The word "open" appears 19 times in Huang's relatively short announcement.

Between the September 3, 2026 announcement and the earliest possible close in the first half of 2027 there are roughly four to ten months.

The commitments enumerated in Huang's announcement concern choice of model, framework, cloud, inference provider and computing platform plus continued multi-accelerator support; the announcement contains no commitment on pricing, rate limits, or which artifacts are presented to developers by default.

In late July Nvidia reported $36 billion in commitments from agreements with AI cloud partners; under these deals Nvidia's commitments shrink when partners sell their capacity to third parties, and Nvidia also partly insures cloud partners against unused capacity.

Nvidia has a market capitalisation of well over $5 trillion and is the most valuable company on Earth.

Nvidia is paying about 1.7 times what Microsoft paid for GitHub, in nominal dollars.

The staff retention stock program of up to $1 billion is roughly 8 percent of the $12.9 billion total.