Skip to main content
Nvidia CEO Jensen Huang and Hugging Face logo, symbolizing the reported $13 billion acquisition talks.

Editorial illustration for Report: Nvidia to Buy AI Model Hub Hugging Face for USD 13 Billion

Nvidia to Acquire Hugging Face for $13B

Report: Nvidia to Buy AI Model Hub Hugging Face for USD 13 Billion

Updated: 4 min read

Nvidia is close to buying Hugging Face for $12.9 billion, according to a report from The Information that cites a person with knowledge of the matter. CNBC has a second source confirming talks are underway. Nothing is signed yet, so the deal could still fall apart, but both sides appear to be pushing toward an agreement.

Hugging Face works like a version of GitHub built specifically for AI models. Researchers and developers use it to search for models, download them, fine-tune them, and upload the resulting variants back to the platform. Founded in 2016, the company has become something close to a default hub for the open-model ecosystem, which makes it a target worth fighting over.

Salesforce reportedly tried to buy Hugging Face as well, and Nvidia isn't a new name on its cap table. Nvidia, Google, and Microsoft have all invested in the company at various points.

For Nvidia, the acquisition would tighten its grip on the AI stack far beyond chips. It also lines up with the company's public bet on open-weight models, a bet it's backed with its own Nemotron model family.

Nvidia is reportedly moving forward to acquire Hugging Face for $12.9 billion. The acquisition could help the hardware giant expand and fortify its deep integration with the wider AI industry.

Why this matters If this deal closes as reported, the company that already controls most of the compute layer for AI would also own the place where a huge share of the field's models get published, downloaded, and fine-tuned. That's worth sitting with. Hugging Face works because it reads as neutral ground, a shared library that any lab, startup, or hobbyist can use regardless of which GPUs they're running on.

Nvidia writing a $12.9 billion check changes that calculus, even if nothing about the platform's mechanics changes on day one. For developers and founders building on Hugging Face, the practical question is whether model hosting, licensing terms, or integration priorities start tilting toward Nvidia's stack over time. For researchers, it's whether "open" repositories stay open when the landlord is also the industry's dominant hardware vendor.

We'd watch how Hugging Face's leadership talks about independence in the coming weeks, and whether rival chipmakers or cloud providers respond by backing alternative repositories. Consolidation at this scale rarely stays quiet for long.

Common Questions Answered

What is the reported acquisition price for Nvidia's purchase of Hugging Face?

According to reports from The Information and confirmed by CNBC, Nvidia is close to acquiring Hugging Face for approximately $12.9 to $13 billion. However, the deal has not been finalized yet, and negotiations are still ongoing between both parties.

How does Hugging Face function as an AI model hub?

Hugging Face operates similarly to GitHub but is specifically designed for AI models, allowing researchers and developers to search for models, download them, fine-tune them, and upload their own contributions. It serves as a shared library accessible to labs, startups, and hobbyists regardless of which GPUs they use.

What concerns does the Nvidia-Hugging Face acquisition raise about industry neutrality?

Hugging Face has been valued as neutral ground where any developer can access and share AI models independently of their hardware choices. Nvidia's $12.9 billion acquisition could change this perception since Nvidia already controls most of the compute layer for AI, potentially giving it significant influence over where the field's models are published and distributed.

Why would this acquisition help Nvidia expand its position in the AI industry?

The acquisition would allow Nvidia to deepen its integration with the broader AI industry by owning both the hardware infrastructure (compute layer) and the primary platform where AI models are published, downloaded, and fine-tuned. This combination would give Nvidia unprecedented control over critical parts of the AI development and deployment ecosystem.

LIVE15:09Google's WikiSkill gives AI agents a persistent memory to avoid past mistakes