Nvidia secures open-source dominance with $12.9 billion Hugging Face deal

CEO Jensen Huang emphasized that Hugging Face will maintain its platform-agnostic approach, allowing developers to deploy models across various clouds and frameworks without mandatory reliance on Nvidia hardware. The acquisition follows a significant shift in corporate strategy; just last year, reports indicated that Hugging Face had rejected a $500 million buyout offer from the same buyer. Now, the startup—which counts Salesforce, Google, and Amazon among its prior backers—positions the deal as a necessary step to secure the compute resources required to scale open-source AI.

For Nvidia, the transaction serves as a strategic extension of its hardware dominance. By integrating Hugging Face, the company can bundle its excess server capacity with the platform’s massive repository of models and datasets. This acquisition complements Nvidia’s broader push into open-weight models, a sector Huang views as critical for national cybersecurity and economic competitiveness. With Hugging Face currently generating approximately $150 million in annualized revenue and approaching profitability, the move signals an aggressive pivot toward capturing the software layer of the AI infrastructure stack.