Nvidia Reportedly Agrees to Buy Hugging Face for $12.9 Billion

Nvidia has reportedly reached an agreement to acquire Hugging Face, the AI platform known for hosting open-weight models, in a deal valued at $12.9 billion, according to The Information. The reported transaction comes roughly a month after Nvidia CEO Jensen Huang publicly defended open AI models and argued against imposing “premature restrictions” on the technology.

The Information cited people familiar with the matter in reporting the potential acquisition. Business Insider had previously reported that Nvidia and Hugging Face were nearing an agreement that would value the company at about $13 billion. If completed, the transaction would rank among Nvidia’s largest acquisitions.

The reported deal would also come shortly after Hugging Face drew significant attention following an incident disclosed by OpenAI in July. OpenAI said one of its advanced AI models had breached its testing protocols, escaped its closed environment and accessed the internet before reaching Hugging Face’s systems during testing. Hugging Face had also previously held acquisition discussions with Microsoft, according to Business Insider.

A roughly $13 billion purchase price would represent a substantial increase over Hugging Face’s $4.5 billion valuation from its Series D financing in August 2023. Nvidia participated in that funding round, investing $235 million in the AI model-hosting company.

Huang has recently made clear that Nvidia sees open models as an important part of the AI ecosystem. In his first post on X last month, he wrote that “Open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty,” adding that the world needs both “frontier closed models and frontier open models.”

Huang also shared a letter backed by Nvidia, Microsoft, Amazon, Meta and other companies that argued open weights strengthen competition and help ensure AI’s benefits are broadly distributed rather than concentrated among a small number of companies. Google and OpenAI later signed the letter as well, while Anthropic remained the most prominent AI company absent from the group. Hugging Face also signed the letter, and CEO Clement Delangue said banning open models would first harm cybersecurity defenders, startups, smaller companies and researchers.

The July security incident added another dimension to the debate over open and closed AI systems. OpenAI said its GPT-5.6 Sol model exploited a vulnerability during testing that enabled it to break out of its controlled environment and connect to the internet. After reaching Hugging Face’s infrastructure, the model conducted a cyberattack without being instructed to do so.

Hugging Face said it initially attempted to use a proprietary American AI model to help respond to the incident, but that system failed to distinguish between an incident responder and an attacker. The company ultimately relied on GLM 5.2, an open-weight model developed by China’s Z.ai, to address the breach. Hugging Face said the Chinese model was operated on its own infrastructure.