Nvidia CEO Says Hugging Face Will ‘Remain an Open Platform for the Entire AI Ecosystem’

Nvidia is officially buying AI resource repository Hugging Face in a whopping $12.9 billion bet on open models. The platform, and the 3 million open models it houses, will now be under Nvidia’s management. More than 200 thousand companies rely on the platform in their AI deployment, Nvidia shared in its announcement of the acquisition, while promising to retain the platform’s open nature. “Hugging Face will remain an open platform for the entire AI ecosystem,” Nvidia CEO Jensen Huang said in the blog post. “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.” Hugging Face has been at the center of the global open model world for some time now, but the name shot to mainstream fame in July when OpenAI admitted that its agents had gone rogue and hacked into the repository’s system. The incident also brought the supposed importance of open models to light after Hugging Face said that it had to resort to a Chinese open-source model to respond to the attack when the closed models wouldn’t work due to cybersecurity restrictions. Silicon Valley giants came out in support of advancing the open model space after the incident. Chief among them was Nvidia, which has long been openly and vocally interested in advancing open models, portraying them as a key way for America to establish global AI dominance. As American AI model providers switch focus to closed models, Chinese AI labs like DeepSeek and Moonshot have established a dominant presence in open models. That presents risks to the American AI industry, Huang has argued, because Chinese AI companies could translate a global reliance on Chinese open models to a reliance on Chinese AI chips. Huang’s fears surrounding Chinese dominance might be spelling out the path forward company leadership has chosen for Nvidia itself. Nvidia is already the leading name in AI hardware, but it increasingly wants to secure its position on the software side as well. Late last year, the company released its latest family of open models, signaling its renewed interest in becoming a bigger player in open models. Nvidia dominating both the hardware and the open model space could help the company ensure that the most seamless option for developers building AI on open models would be to keep it in-house and use Nvidia’s own chips to run them. The strategy is especially crucial at a time when most of Nvidia’s top chip customers building the frontier closed models, such as OpenAI and Google, are building their own in-house chips instead, positioning the now-collaborators as future competitors. Hugging Face and its wide reach in the developer world seem to fit perfectly within this strategy. The chip giant even wanted to invest $500 million in Hugging Face earlier this year, but Hugging Face ultimately refused, telling the Financial Times that it doesn’t want a single dominant investor on board swaying decisions. But Hugging Face seems to have changed its mind on that rather quickly. Reports claimed early last month that it was looking to get acquired, followed by a report from The Information that claimed Nvidia had stepped up to be the buyer. At the time of Nvidia’s rejected investment offer, the FT put Hugging Face’s valuation at $7 billion, far below the $12.9 billion it eventually sold for roughly 8 months later. The number is so big that it marks Nvidia’s second-largest acquisition, just after the $20 billion it shelled out in December to acquire the flashy AI inference chips startup Groq (not to be confused with Musk’s Grok). As company executives pointed out to the FT back in January, Hugging Face’s strength in open models comes from its neutrality. The platform’s acquisition by such a major name with a substantial stake in the game could make the developers who use the platform and value that reputation uneasy. But Huang promises to honor Hugging Face’s openness. “Together, we will scale Hugging Face’s platform, strengthen its infrastructure and expand access to AI for developers and institutions worldwide,” Huang said in the blog post. “NVIDIA’s infrastructure, engineering and global reach can help improve platform reliability, safety, model evaluation, inference and deployment capabilities, while preserving the open ecosystem that made Hugging Face foundational.”