On September 3, 2026, NVIDIA announced an agreement to acquire Hugging Face for $12.93 billion. Hugging Face is the default public repository for open AI work: NVIDIA’s announcement cites more than 18 million developers, researchers and creators on the platform, sharing more than 3 million models, 500,000 datasets and 1 million applications, with more than 200,000 companies using it to discover, evaluate, customize and deploy AI.
The commitments attached to the deal are as notable as the price. NVIDIA stated that Hugging Face will remain an open platform for the entire AI ecosystem, that developers keep their choice of models, frameworks, clouds and inference providers, and explicitly that NVIDIA compute will not be required to build on or deploy through Hugging Face. Jensen Huang described the intent as to “scale Hugging Face’s platform, strengthen its infrastructure and expand access to AI for developers and institutions worldwide,” and noted that co-founder Clement Delangue approached him as Hugging Face considered its next chapter.
The strategic logic runs down the stack rather than across it. NVIDIA already owns the accelerator, the interconnect and increasingly the rack; Hugging Face is where the artifacts that run on that hardware are published, benchmarked and pulled. Owning the distribution layer for open weights puts NVIDIA at the point where a developer chooses a model, which is upstream of where they choose a chip.
For technical leaders, the open-platform pledge is the thing to watch rather than take on faith. A very large share of enterprise AI supply chains now pulls artifacts from a single hub; that hub having a hardware vendor as its owner is a governance question worth writing into vendor reviews, alongside the reminder that this same platform was the subject of a serious intrusion earlier in 2026.