NVIDIA is acquiring Hugging Face for $12.93 billion, a move that puts the largest AI hardware company in charge of the open model repository that more than 18 million developers rely on. The platform currently hosts over three million models, 500,000 datasets, and one million applications, with commercial adoption crossing 200,000 companies. That’s not a community. That’s critical infrastructure.
Jensen Huang was quick to address the obvious fear — that NVIDIA would turn Hugging Face into a CUDA-only showroom. “NVIDIA compute will not be required to build on or deploy through Hugging Face,” he said, adding that developers will keep full control over their choice of models, frameworks, clouds, inference providers, and hardware. The platform will continue supporting alternative accelerators and multi-cloud architectures. Whether that promise survives contact with quarterly earnings pressure is the question everyone should be asking.
Hugging Face co-founder and CEO Julien Chaumond framed the deal in stark terms: open-source AI could “become less relevant in the coming years if the big closed labs run away with it” or become “the foundational fabric of the next phase of human civilisation.” He said NVIDIA was “the only partner we truly considered,” citing Huang’s public defense of open weights earlier this summer. That’s a notable endorsement from someone who spent a decade building an independent platform.
The acquisition gives NVIDIA something it has never really had: a direct, trusted channel to the developer community that actually builds and fine-tunes models — not just the enterprises buying DGX systems. NVIDIA is already the largest contributor of open models and datasets to Hugging Face, with more than 500 open models and 250 open datasets. The brand stays independent, the team stays in place, and the infrastructure investment is supposed to improve reliability, safety controls, and deployment pipelines. But when the largest compute vendor owns the primary distribution channel for open models, the line between neutrality and strategic advantage gets very thin, very fast.
