Nvidia has agreed to acquire Hugging Face, the open-source AI model repository, for $12.9 billion, according to a person with knowledge of the deal.
The transaction values Hugging Face at nearly double the $7 billion offer the company rejected in 2025, and marks a sharp escalation from its $4.5 billion valuation in 2023. Hugging Face hosts machine learning models and datasets used by researchers and developers across the industry; the platform has become central infrastructure for the open-source AI ecosystem.
The acquisition places a major hardware maker in direct control of a primary distribution channel for AI models that compete with Nvidia's own software offerings. Hugging Face's repository hosts thousands of open-weight models from Meta, Mistral, and other AI builders; researchers and startups use the platform to download, fine-tune, and deploy models without licensing constraints. The deal gives Nvidia ownership of that traffic and the usage data flowing through it.
Hugging Face raised $4.5 billion at its last disclosed valuation in 2023, according to previous filings. The company was founded in 2016 and operates on a freemium model, with most model downloads free and commercial features as paid add-ons. Its user base spans academic researchers, enterprise teams, and AI startups.
Nvidia's move follows an 18-month period of consolidation in AI infrastructure. The company has spent heavily on acquisitions in networking, chip design, and software layers as competition in AI accelerators intensifies from AMD, Intel, and custom silicon makers. The Hugging Face deal, if closed, would be one of Nvidia's largest acquisitions by value.
The deal structure and expected close date have not been publicly disclosed. Nvidia and Hugging Face have not issued a joint announcement as of publication.
The $12.9 billion valuation exceeds Hugging Face's previous rejection of a $7 billion offer by 84 percent, a jump driven by the surge in large language model deployment and Nvidia's deepening focus on controlling the full AI software stack. The timing of the acquisition aligns with intensifying pressure on open-source model distribution platforms as enterprise adoption of AI accelerates.
