Nvidia Buys Hugging Face for $12.9 Billion
By @sharedot · · 7 pages
- AI
- Nvidia
- Hugging Face
- M And A
- Open Source
Nvidia is buying open-source AI hub Hugging Face for $12.9 billion, promising the platform stays open.
What happened
Nvidia has agreed to acquire Hugging Face, the open-source platform hosting more than 3 million models, 500,000 datasets and 1 million applications, for $12.9 billion.
The deal was confirmed on Thursday, September 3, after initial reports, and includes $1 billion for a Hugging Face employee retention plan, according to CEO Jensen Huang's comments to CNBC.
Hugging Face was last officially valued at $4.5 billion after a 2023 funding round that Nvidia itself participated in, per a Financial Times report cited by The Indian Express.
The acquisition lands in the same week that Anthropic, OpenAI, Meta and Google all shipped model updates within days of one another, making it the capstone of one of AI's busiest news cycles yet.
Why it is surprising
Hugging Face has long positioned itself as the neutral, open alternative to the walled gardens of OpenAI and Google — essentially the GitHub of machine learning. Selling to the company that dominates AI hardware is a striking reversal of that positioning.
Scripps News notes the platform was just hacked in July, in an incident OpenAI acknowledged its system caused, making Nvidia's multibillion-dollar rescue an unexpected twist.
Gartner analyst Arun Chandrasekaran told The Indian Express the deal proves 'the next battleground in AI is who owns the developer rather than the best model.'
The $12.93 billion figure also makes this one of Nvidia's largest acquisitions, second only to its $20 billion Groq assets deal, according to The Indian Express.
The evidence behind the move
Huang's blog post, cited by Scripps News, framed the deal as a bet on open-source AI: 'Open models let startups, businesses, universities and public institutions build on advanced capabilities without training every model from scratch.'
He said Hugging Face will remain open, supporting multi-cloud and multi-accelerator deployment, and that Nvidia compute will not be required to build or deploy through the platform.
The scale is substantial: more than 18 million developers and over 200,000 companies use the platform, and Nvidia has itself uploaded more than 500 models and 250 open datasets there.
The Indian Express reports Nvidia hopes the deal could help Hugging Face grow its user base from 18 million developers to 100 million.
The stakes for open-source AI
The core question is whether Hugging Face's neutrality can survive.
The Indian Express reports rivals like AMD and Broadcom may become reluctant to rely on a platform owned by their dominant competitor, and that commenters noted Nvidia could still optimize the platform so models run fastest on its own hardware.
Chandrasekaran warned that 'true neutrality may not be possible' under a major player's ownership. Antitrust scrutiny looms too: Nvidia holds more than half of the GPU market for AI data centers and has faced probes from the US Justice Department, France and China, The Indian Express reports.
For enterprises, consolidating infrastructure ownership on top of relentless release cycles makes migration decisions even harder, The Tech Buzz notes.
What comes next
Whether the deal clears regulators may depend on geography: The Indian Express suggests it may pass US scrutiny given the administration's AI focus, but European approval is uncertain.
Nvidia's bet, per Huang, is that businesses will increasingly turn to open-source models rather than pay for proprietary services from OpenAI and Anthropic — a hedge West Monroe's Bret Greenstein, quoted by Scripps News, called 'a strong strategic hedge against the different ways AI could evolve.'
The Indian Express also frames the deal as strengthening US open-source AI against a wave of cheap Chinese open-weight models like Moonshot AI's Kimi K3. Watch how Hugging Face's community reacts to life under Nvidia's roof, and whether any lab slows its release pace to pitch fewer, better updates.