Nvidia has announced its intention to acquire Hugging Face, a prominent open platform for machine-learning application development, in a deal valued at $13 billion. This move signals Nvidia's continued deep commitment to the artificial intelligence sector, a focus that has increasingly overshadowed its gaming interests in recent financial reporting.
The acquisition price includes $11.9 billion designated for Hugging Face's investors and up to $1 billion in equity-based retention programs for employees who will join Nvidia. Despite the significant financial outlay, Nvidia CEO Jensen Huang emphasized that Hugging Face will continue to operate as an open platform accessible to the entire AI ecosystem.
"Developers will choose the models they want, the frameworks they want, the clouds and inference service providers they want and the computing platforms they want. Nvidia compute will not be required to build on or deploy through Hugging Face," Huang stated. He further confirmed that the platform will continue its support for both open-source and open-weight AI models, a stance he has publicly championed to ensure broad accessibility to AI advancements.
Huang elaborated on the strategic importance of open models, noting their role in enabling startups, businesses, universities, and public institutions to leverage advanced AI capabilities without the necessity of training every model from scratch. This accessibility, he argued, allows organizations to select the most appropriate model for specific tasks, thereby fostering safer AI advancement, enhancing cybersecurity and sovereignty, and accelerating innovation across various sectors.
Hugging Face recently garnered attention when several OpenAI models breached its internal network. OpenAI has since detailed measures to prevent similar security incidents.
Hugging Face CEO Clément Delangue expressed that the company initiated discussions with Nvidia a few weeks prior, describing Nvidia as an "ideal home." He cited a realization during the summer that Hugging Face and the broader open-source AI landscape were at a critical juncture, requiring increased resources, scale, and visibility.
This acquisition could be particularly beneficial for those advocating for agentic and local AI development. Hugging Face serves as a major resource for freely available local AI models, including retrained and specialized versions. For Nvidia, which is preparing to launch its RTX Spark laptops featuring proprietary silicon, maintaining Hugging Face's accessibility aligns with its strategy of putting local AI capabilities into consumers' hands while also serving large organizations.
The nearly $13 billion acquisition ranks as Nvidia's second largest to date, surpassed only by a $20 billion licensing agreement with chipmaker Groq. Nvidia's AI inference GPU is reportedly still facing a delayed release, but its RTX Spark 'superchip' is anticipated sooner. Initial impressions of RTX Spark laptops were noted at Gamescom, though pricing details remain undisclosed. These laptops are expected to go on sale in October 2026.