Nvidia to acquire open-source AI platform Hugging Face for $12.93 billion

Nvidia has officially announced an agreement to acquire Hugging Face, the central repository and collaboration platform for open-source machine learning, for $12.93 billion. The transaction, which is expected to close in the first half of 2027 pending regulatory approval, marks a significant expansion of Nvidia's footprint in the open-weight AI model space.

Hugging Face currently serves as a foundational hub for the AI community, hosting over 3 million models, 1 million applications, and 500,000 datasets. The platform supports more than 18 million developers and 200,000 enterprise users. According to Hugging Face CEO Clément Delangue, the company reached an inflection point where the resources required to scale and support the growing open-source community exceeded its independent capacity.

Nvidia CEO Jensen Huang stated that the acquisition aims to provide Hugging Face with the infrastructure, engineering resources, and global reach necessary to scale its operations. Despite the acquisition, both companies have committed to maintaining Hugging Face as an open platform. Developers will retain the ability to choose their preferred frameworks, cloud providers, and computing hardware, with Nvidia explicitly noting that its own proprietary compute will not be a requirement for using the platform.

The deal follows Nvidia's long-standing involvement in the open-source community, where it has contributed over 500 models and 250 datasets to the Hugging Face hub. Huang emphasized that the company remains committed to supporting a diverse ecosystem of open models, which he argues are essential for innovation across industries ranging from robotics and biomedical research to cybersecurity.

The acquisition comes as open-source AI models see increased adoption, with recent data indicating a surge in contributions from various global labs. As the industry navigates the divide between proprietary APIs and open-source alternatives, Nvidia’s investment positions it to influence both the hardware infrastructure powering AI and the platforms where developers build and deploy their models.