NVIDIA to acquire Hugging Face for $12.93 billion in largest AI infrastructure deal

NVIDIA has agreed to acquire Hugging Face for approximately $12.93 billion — about $11.9 billion in cash plus up to $1 billion in retained equity — in what is being described as the largest AI-infrastructure acquisition to date. Hugging Face is the de facto commons of open machine learning, hosting more than 3 million models, 500,000 datasets, and a community of 18M+ developers. Jensen Huang publicly pledged the platform will remain open, multi-cloud, and multi-accelerator, and NVIDIA is positioning the deal as expanding the 'front door' through which developers reach the compute-hungry workloads that ultimately run on its GPUs.

Strategically, the logic is a demand funnel: open models drive experimentation, experimentation drives training and inference, and both drive chip sales. At IBC, NVIDIA paired the news with real-time broadcast AI tooling including a Synthetic Video Detector NIM microservice it claims hits 99.3% accuracy on AI-generated footage.

The developer reaction is the story's live wire. On Reddit and HN, many were cautiously optimistic but wary: 'the open-model commons is being recapitalized by the chip company that sells the GPUs.' The neutrality of a vendor-owned hub — especially one whose owner has an obvious interest in steering workloads toward CUDA — is the core anxiety, only partly offset by Huang's multi-accelerator pledge. Ironically, Hugging Face CEO Clement Delangue has spent the week arguing for '100x more transparency in AI.' What to watch next: regulatory scrutiny in the US and EU (the deal isn't expected to close until H1 2027), whether NVIDIA commits contractually to accelerator neutrality, and whether rival hubs or forks emerge as insurance.