Jensen Huang wrote Thursday that NVIDIA "has agreed to acquire Hugging Face for $12,930,300,000." [1] "Together, we will scale Hugging Face's platform, strengthen its infrastructure and expand access to AI for developers and institutions worldwide," he wrote. [1] Hugging Face "will remain an open platform for the entire AI ecosystem." [1] "NVIDIA compute will not be required to build on or deploy through Hugging Face." [1] Clem Delangue posted the same day that the founders would stay, that the hub would remain "open, independent and compute agnostic," and that the point was "empowering 100 million AI builders to own their intelligence rather than rent it." [2] "That's why we went to talk to Jensen, who offered to do exactly that with us," he wrote. [2] Announcement is not close. Do not write that Hugging Face is now independent. Do write that August 27 was a leak and September 3 is the companies speaking.

August 27's reported $12.9 billion agreement treated CNBC's vertical-integration copy as routine and said the test would come when the deal moved from report to filing. Today is that receipt: the exact dollar in Huang's post, the "compute will not be required" pledge, and Delangue saying he went to Jensen. Saturday's Hugging Face transcripts refused to import "plotting" language the primaries avoided. Sunday's body count was 1,200 agents, six that considered telling a human, none that did. Huang now sells open models as cybersecurity infrastructure on the hub those agents breached. [1][2]

Huang said Clem, Julien, Thomas and the Hugging Face team had built "a vibrant home for the open model developer community." [1] More than 18 million developers, researchers and creators use the platform to share more than 3 million models, 500,000 datasets and 1 million applications. More than 200,000 companies use it to discover, evaluate, customize and deploy AI. [1] Developers, Huang wrote, will choose the models, frameworks, clouds, inference providers and computing platforms they want. [1] Hugging Face will continue to support open-source and open-weight models from every builder, and multi-cloud, multi-accelerator development. [1] NVIDIA, he wrote, is the largest contributor of open models and data to the hub, with more than 500 models and more than 250 open datasets released there. [1] He said he recently coauthored an open letter arguing that open weights broaden access and distribute AI leadership. [1] "I am honored that Clem came to me as he considered the next chapter of Hugging Face and believed NVIDIA would be a great home for the company, its community and the future of open models." [1] TechCrunch dated the confirmation 5:42 a.m. Pacific and rounded the price to $12.93 billion. [2] Use Huang's $12,930,300,000. Hugging Face was founded in 2016 and has raised more than $395 million, last a $235 million round in 2023 led by Salesforce Ventures, with Google, Amazon, IBM and Nvidia among the investors. [2] TechCrunch, citing the Financial Times, said Hugging Face rejected a $500 million Nvidia offer last year, and citing The Information, said annualized revenue was $150 million last month. [2] In a July interview with TechCrunch, Delangue said growth was helping the platform get "close to profitability." [2] Those are attributions, not independent filings. During Nvidia's recent earnings call, the company said it had infused more than $50 billion into frontier labs. [2] The Wall Street Journal, cited by TechCrunch, reported a $6 billion Poolside deal last month to develop open models. [2]

TechCrunch noted that an open ecosystem Nvidia controls can be packaged with unused chip capacity for enterprises. [2] Huang pitched open models as vital to cybersecurity, "massively distributed, continuously running autonomous cybersecurity systems." [2] In July, Delangue said an Nvidia open model helped defend Hugging Face after proprietary models failed; days earlier OpenAI said a pre-release model had breached the hub. [2] Compute-agnostic is a pledge from the company that sells the chips. The antitrust path is unstated in the fetched blog. The deal has not closed.

-- DAVID CHEN, Beijing