Hugging Face’s Monetization Chief Jeff Boudier Isn’t Interested in Chasing Money

Hugging Face, the A.I. startup known for its GitHub-like hosting platform used by millions of developers and businesses, is pushing back against Silicon Valley’s obsession with relentless fundraising and hyper growth. The New York City-based company, which hasn’t raised venture capital in more than two years, is in no rush to seek new funding or pursue lucrative revenue opportunities like introducing ads, said Jeff Boudier, the company’s head of monetization.

“In a funny way, we’re taking a contrarian approach,” Boudier, who has led product and growth at Hugging Face since 2020, told Observer. “We’re not famous for making money. We’re famous for making A.I. more accessible and easier to use.”

A former business development executive at GoPro, Boudier joined Hugging Face as its first “business-focused” employee, in his words, tasked with building up the company’s business model. Five years into his tenure, Hugging Face is now either net profitable or making investments, depending on the quarter, he said.

Hugging Face was founded a decade ago by French entrepreneurs Clément Delangue, Julien Chaumond and Thomas Wolf. It started out building a chatbot but later gained prominence as a platform where researchers and developers share models and datasets, with a strong emphasis on open-source code. Its offerings extend beyond large language models to include speech, image and video. Today, the platform hosts 2.5 million models, 700,000 datasets and has 13 million users worldwide.

Hugging Face hasn’t raised funds in more than two years. Its last round, a $235 million Series D in 2023, valued the company at $4.5 billion—a figure impressive at the time but modest by today’s A.I. standards. Last year, the company reportedly turned down a $500 million investment from Nvidia, according to the Financial Times. The publication reported that Hugging Face rejected the offer to avoid taking on a single, dominant investor. Boudier said the company still has money in the bank from the 2023 round and that funding isn’t a “priority” right now. Instead, he’s focused on building a financially sustainable platform.

How does Hugging Face make money?

Hugging Face operates a “freemium” business model, where most users access its services for free and a small portion—roughly 3 percent to 5 percent of users—pay for high usage or enterprise collaboration, Boudier said. Paying clients include major tech players like Meta, Google, Microsoft, Apple, OpenAI and Anthropic.

In recent months, Hugging Face has also been experimenting with a new, albeit small, revenue stream: robots. At the end of last year, it began shipping the Reachy Mini, a small, desktop-sized robot that starts at $299 and allows users to experiment with various A.I. applications. More than 5,000 units have been sold. The robot made an appearance during Nvidia CEO Jensen Huang’s CES keynote in January to demonstrate the capabilities of Nvidia models. “It’s a successful experiment, and we’re looking to scale it this year,” said Boudier.

Hugging Face is investing heavily in robotics, betting that demand for on-device A.I. will continue soaring, Boudier noted. Production of the Reachy Mini was made possible by Hugging Face’s acquisition of Pollen Robotics, a developer of open-source humanoid robots, in 2025. The previous year, Hugging Face launched the LeRobot library, which targets software for robotics systems.

“No” to ads in chatbots

In-app advertising has sparked a debate across the A.I. community in recent weeks. OpenAI recently began introducing ads in ChatGPT. The move that was condemned by rival Anthropic. Hugging Face has faced the same question but decided to stay ad-free. A few years ago, Hugging Face was approached by ad networks interested in integrating ads into its chat app, Hugging Chat.

“We didn’t consider at all these proposals,” said Boudier. “In my view, it was just completely off what we’re trying to do and our values and business model.”

Boudier said the introduction of ads will only further complicate the delicate balance of trust between users and A.I. tools. “It’s interesting to see how it plays out.”