Author | Wildcard

Editor | Jingyu

12.9 billion dollars, Nvidia is set to acquire Hugging Face.

On August 26 local time, foreign media broke the news first; just two days earlier, it was only reported that Hugging Face was "exploring a sale," and now it has turned into "the two parties have reached an agreement." The speed is too fast to digest.

This is not an ordinary acquisition of an AI company; it's the first time the king of chips has reached out to grasp the "distribution rights" in the open-source AI world.

Hugging Face is called the "GitHub of AI," hosting over 3 million model repositories and serving 13 million developers. Meta's Llama, Alibaba's Qwen, and Mistral's open-source models all reside on this platform. It is the de facto "default distribution layer" of the entire open-source AI ecosystem.

Meanwhile, Nvidia has already monopolized the GPU hardware for training and inference.

Now, it aims to own the pathway of model circulation.

01 Rolling in AI

First, let's talk about money. Hugging Face's last funding round was the D round in August 2023, led by Salesforce, which invested $235 million at a valuation of $4.5 billion. Google, Nvidia, Amazon, IBM, Intel, AMD, and Qualcomm all participated in that round. Three years later, the price of $12.9 billion is roughly three times the previous valuation.

At first glance, tripling seems exaggerated. However, in the context of the valuation bubble in the AI industry, this premium actually seems restrained. Perplexity is negotiating funding with a valuation exceeding $30 billion, and the valuations of Anthropic and OpenAI are two orders of magnitude higher than their revenues.

Hugging Face's ARR (Annual Recurring Revenue) is estimated by Sacra to reach $150 million by August 2026, compared to about $30 million in 2023, which is a fivefold increase. With 50,000 enterprise customers and 769 employees, CEO Clément Delangue recently mentioned on a podcast that the company is "close to profitability."

But the $12.9 billion price is not buying Hugging Face's revenue; it is purchasing its position in the AI world.

This logic is closer to IBM's $34 billion acquisition of Red Hat in 2019—strategic buyers pay for positioning in the developer ecosystem rather than for short-term cash flow.

02 Nvidia, Greater Ambitions

If this acquisition is only understood as "a chip company buying a model hosting platform," that would be too simplistic.

Over the past year, Nvidia's investment in open-source AI has been surprisingly substantial.

At the GTC conference in March this year, Jensen Huang announced the establishment of the "Nemotron Alliance," a global collaborative organization consisting of eight AI labs, including Mistral AI, Perplexity, Cursor, LangChain, etc., aimed at jointly developing state-of-the-art open-source foundational models. Nvidia committed $26 billion in investments over five years, disclosed through SEC filings.

This is the largest funding commitment to open-source AI in history.

The first model developed by the alliance will form the foundation of the Nemotron 4 series. According to foreign media reports in early August, the largest version of Nemotron 4 is expected to reach a trillion parameters, directly competing with the world's strongest open-source and closed-source models. Before this, Nvidia had already released the Nemotron 3 series, including the Nemotron 3 Super announced in March (120 billion parameters, 12 billion activation parameters) and the just-released Nemotron 3.5 Lightning in August, a lightweight open-source model that can run on a single GPU.

Jensen Huang personally defended open-source models.

In July this year, he posted his first message on X, publicly supporting open-source AI, coinciding with the surge in national security discussions in the U.S. over the Kimi K3 model from Moonshot AI. His stance was clear: "Free AI is good for hardware. Free AI is good for chips."

In plain terms, Nvidia's underlying logic in promoting open-source is exactly the same as Google's push for Android—making software free while using the software ecosystem to drive hardware sales. Every developer using open-source models to create products will need Nvidia's GPUs for training and inference. The more models there are, the more users there are, and the greater the demand for chips.

Now, by acquiring Hugging Face, Nvidia will not just be "providing hardware to train open-source models," but will directly own the place where the open-source models "live."

03 Embodied Intelligence, a Larger Chess Game

However, Nvidia's appetite goes beyond language models and code generation.

This year, also at GTC, Jensen Huang unveiled Isaac GR00T N1, which claims to be the world's first open-source general humanoid robot foundational model. In June's GTC Taipei, he presented the Isaac GR00T reference humanoid robot—a complete open-source hardware reference design based on the Yu Shu H2 Plus humanoid chassis, Sharpa five-finger dexterous hand, Jetson Thor computing platform, and the entire Isaac GR00T software stack. Stanford, ETH Zurich, Allen AI Research Institute, and UCSD are among the first collaborating institutions.

At the same time, Cosmos 3, released at the end of May, is Nvidia's first fully open-source "multi-modal model," designed specifically for "physical AI"—that is, controlling robots, autonomous vehicles, and other machines operating in the physical world. The autonomous driving inference model Alpamayo, released at CES in January, is also open-source.

Looking at these points collectively, Nvidia is building a complete open-source ecosystem for "physical AI." From the world model (Cosmos) to the humanoid foundational model (GR00T N1) to the simulation engine (Isaac Sim), hardware reference design, and even edge computing chips (Jetson Thor), the full stack is covered, all open-source or open.

Jensen Huang stated in Taipei: "Humanoid robots will bring physical AI to the world's largest industry, unlocking a multi-trillion-dollar economic opportunity."

Hugging Face, conveniently, is the default platform for robot researchers worldwide to share models, datasets, and tools. When Nvidia owns this distribution layer, its embodied intelligence open-source ecosystem transforms from "bulk" to a "full closed-loop."

Training on Nvidia's GPUs, models published on Nvidia-owned platforms, running on Nvidia's chips, controlling robots powered by Nvidia's computing capabilities.

04 Doubts About the "Neutral Platform"

The most concerning question for everyone is the "neutrality" of Hugging Face.

As early as January this year, there were reports that Hugging Face rejected Nvidia's $5 billion investment, primarily because the founding team was concerned that being too closely tied to a certain chip giant would harm the platform's neutrality among competitors like AMD and Intel.

But seven months later, they chose to accept the $12.9 billion acquisition.

Concerns from the developer community are valid. Some users on X stated frankly: "One way to suppress the open-source path is to let Hugging Face be acquired by a big company." The platform hosts a large number of open-source models from Chinese companies, with DeepSeek and Alibaba's Qwen already being the most popular options on Hugging Face by download count in multiple categories. Can a platform belonging to a U.S. chip giant still act as a globally neutral AI model distribution center?

History provides two references. In 2018, Microsoft acquired GitHub for $7.5 billion, causing developer panic and mass exits, but Microsoft ultimately maintained GitHub's independent operations and open ecosystem, making it even better. After IBM acquired Red Hat for $34 billion in 2019, it also promised to preserve its open-source culture. These acquisitions ultimately proved that "incorporating" open-source platforms does not necessarily mean "closing off."

But Nvidia's situation is fundamentally different.

When Microsoft acquired GitHub, Microsoft was not a monopolist in the developer tools market; when IBM acquired Red Hat, IBM was not the absolute dominant player in the enterprise computing market.

In contrast, Nvidia holds more than 80% market share in the AI GPU market.

A player that monopolizes hardware while also owning the largest distribution platform for open-source software will inevitably be closely scrutinized by regulatory agencies.

The proportions of cash and stock, arrangements for retaining core R&D teams, and methods for continuing the platform's open-source commitments have yet to be disclosed. These details will determine whether this acquisition is a normal expansion of the ecosystem or a vertical integration causing industry upheaval.

From the $26 billion investment to build the Nemotron alliance to the trillion-parameter model, from the GR00T robot to the Cosmos world model, and now to acquiring Hugging Face for $12.9 billion—Nvidia's actions over the past six months have drawn a clear line.

It is no longer satisfied with just selling shovels to those mining for gold; it wants to own the gold mine itself. Or more accurately, it wants to own the entire supply chain from the mine to the distribution market.

What does this mean for the AI industry? When the most powerful AI infrastructure company begins to simultaneously control hardware, models, distribution platforms, and robot development stacks, the meaning of the words "open-source" may need to be redefined.