AI
Nvidia buys Hugging Face for $12.9B, pledges to keep AI platform open and hardware-neutral
The chipmaker's acquisition of the major open-source AI model hub comes with explicit commitments to maintain vendor independence, though skeptics point to Microsoft's GitHub playbook as a cautionary tale.
The New Stack
Nvidia has secured Hugging Face, a central repository for open-source AI models, in a $12.9 billion acquisition that unites the world's most valuable company with one of the AI ecosystem's most critical neutral platforms. The deal sparked immediate concerns about whether the hub might eventually become a vehicle for steering developers toward Nvidia hardware—a worry the chipmaker is working hard to defuse.
Nvidia's neutrality pledge#
Jensen Huang, Nvidia's CEO, has laid out a series of commitments intended to preserve Hugging Face as a genuinely open platform. In Nvidia's official announcement, he states that Hugging Face will support developers across all clouds and accelerators, and crucially, that "Nvidia compute will not be required to build on or deploy through Hugging Face." The word "open" recurs 19 times throughout Huang's relatively brief statement, signaling how central these assurances are to the company's framing of the deal.
Huang also referenced a recent letter on open-weight models he co-signed with AI leaders across the industry, including Hugging Face executives, emphasizing that open-weight systems broaden AI access and foster meaningful competition. Nvidia itself has recently invested in open-source AI through projects like its Nemotron models and efforts to make large models runnable on local hardware.
Why Hugging Face accepted the deal#
Clément Delangue, Hugging Face's co-founder and CEO, framed the acquisition as a necessary step to scale open-source AI infrastructure. He noted that the platform has demonstrated itself as a viable complement to proprietary APIs, but reaching the next level of adoption would demand substantially greater computational resources, organizational support, and public presence. Delangue emphasized that Nvidia has committed to maintaining the platform's independence, compute neutrality, and the continuity of the existing leadership and team.
The deal represents a milestone in the push to make open-source the standard approach to building AI systems. Delangue outlined an ambition to empower 100 million developers to "own their intelligence rather than rent it"—a vision that appears to align with Nvidia's own recent moves into open-source model development.
