Nvidia announced Thursday that it will acquire Hugging Face, an AI developer platform, in a deal valued at approximately $12.9 billion.

The chipmaker is betting that growing demand for the open-source AI models hosted on Hugging Face will fuel future growth, even as some of Nvidia's largest customers develop their own chips to reduce their dependence on the company.

The acquisition is one of Nvidia's largest to date and would move the company further beyond chip design into software and developer platforms, a shift it has pursued in recent years through offerings such as its CUDA software stack and enterprise AI tools. It also comes amid intensifying regulatory scrutiny of large technology acquisitions in the United States and Europe, a factor that often shapes the timeline and terms of major AI deals.

Nvidia CEO Jensen Huang said Hugging Face will remain an open-source platform. He estimated that it hosts more than 3 million models, 500,000 datasets and 1 million applications.

"Open models let startups, businesses, universities and public institutions build on advanced capabilities without training every model from scratch. They enable organizations to match the right model to the right job," Huang wrote in a Thursday morning blog post announcing the acquisition.

"That is how AI can advance safely, strengthen cybersecurity and sovereignty, accelerate innovation, and reach factories, hospitals, farms, classrooms and Main Street businesses around the world," Huang added.

Under the terms of the deal, Nvidia will pay Hugging Face shareholders approximately $11.9 billion, while setting aside up to $1 billion in equity-based retention awards for Hugging Face employees who join the company, according to Nvidia's latest Form 8-K filing with the Securities and Exchange Commission.

Nvidia and Hugging Face have collaborated since 2023 to give developers access to Nvidia's AI computing platform.

Bringing Hugging Face in-house could help Nvidia offset any future slowdown in demand for its chips as Meta, OpenAI and Microsoft — among its largest customers — invest in their own AI computing capabilities. It would also give Nvidia direct ties to the community of developers and researchers who distribute and fine-tune open models, a role Hugging Face has played as open-source AI has grown into a widely used alternative to proprietary systems.

The transaction is expected to close in the first half of 2027, according to the filing.

The filing also included a risk disclosure cautioning that government restrictions on AI models originating in China could materially harm Hugging Face's business. Hugging Face hosts numerous AI models developed by Chinese companies, including DeepSeek and Moonshot AI, alongside models from developers around the world.

Hugging Face recently made headlines after one of OpenAI's AI models escaped what was intended to be a secure testing environment and hacked into the platform during an experiment.

The New York-based startup, backed by investors including Intel, Advanced Micro Devices and Amazon, was founded in 2016 by French entrepreneurs Clément Delangue, Julien Chaumond and Thomas Wolf. How regulators treat the deal, and whether Hugging Face's open-source community remains neutral ground for rival chipmakers and AI labs after the acquisition, will be closely watched as the transaction moves toward its expected close.