Nvidia has agreed to purchase Hugging Face for nearly $13 billion. The deal, which involves the platform that hosts open-source datasets and developer tools, was confirmed on Thursday morning after months of speculation.
The agreement shows Nvidia is investing heavily in the software layer of generative AI, not just the hardware. The chipmaker already distributes a suite of open-weight models called Nemotron. In the announcement, Nvidia promised to keep Hugging Face’s existing open standards intact, positioning itself as a primary resource for developers seeking nonproprietary tools.
The move reinforces Nvidia’s view that open-weight models are superior to closed systems like those from OpenAI and Anthropic. Earlier this month, the company gathered more than 80 firms to sign a letter asking the US government to protect open-weight AI. Nvidia also launched the Shared AI Findings Exchange, an initiative designed to let tech companies confidentially report incidents and publish recommendations to reduce systemic risk.
“Open models let startups, businesses, universities and public institutions build on advanced capabilities without training every model from scratch,” Jensen Huang said in the press release. “AI advances faster when people can build together.”
Nvidia’s support for open-source software might appear contradictory given that CUDA, the proprietary software for its GPUs, is closed. However, as frontier labs and hyperscalers like Amazon and Meta attempt to build custom chips, Nvidia has shifted its strategy to sell high-performance CPUs and expand its software portfolio to retain developers.
Business Insider first reported that the two companies were discussing a merger. The Information later confirmed that a deal had been reached.
History of the platform
Hugging Face was founded ten years ago by three French entrepreneurs in New York City. Their initial plan to launch an AI companion app failed. Instead, the team focused on making complex natural language processing tools easier for developers to use. The company eventually became a hub where users can freely share source code, datasets, and large language models.
“Open-source AI is at an inflection point,” Clément Delangue wrote on X. “Thanks to the community, we’ve shown that it can be a complement, and even an alternative, to closed-source APIs. But for it to happen at larger scale, it needs more compute, more support, more collaboration and more visibility.”
Delangue noted that the platform recently faced a security breach involving AI agents from OpenAI. The incident highlighted concerns about the safety and manageability of autonomous agents promoted by major tech firms.
PitchBook data shows Hugging Face had raised nearly $400 million in venture capital by December 2025. Investors included Betaworks, Lux Capital, Sequoia Ventures, Coatue, and Addition. Notable individual backers are Greg Brockman, Richard Socher, and Kevin Durant.
During an interview with CNBC in 2023, Delangue predicted the number of developers working with AI models would explode. He estimated that within five years, there would be around 100 million AI builders. He believed that if all of them used Hugging Face daily, the company would be in a strong position.
Nvidia’s decision to acquire the platform suggests it agrees with that assessment.
