Monday, August 3, 2026

AI & Models

Hugging Face CEO says high costs drive companies to open source

Hugging Face CEO Clem Delangue says rising scaling costs are driving companies away from proprietary frontier APIs toward open-source AI models.

Open source AI is experiencing a boom, according to Hugging Face CEO Clem Delangue. In recent years, the company has grown into something like a GitHub for AI, a platform where AI builders share and download open models and datasets. That growth reaches deep into the corporate world: roughly half the Fortune 500 — the annual list of the largest US corporations by revenue — now use Hugging Face.

Delangue says he has watched the same pattern play out again and again across the companies that use the platform: enterprises typically begin by renting frontier APIs — proprietary, leading-edge AI models rented from their developers. But as those companies scale up their usage, the costs of continuing to rent access to those models climb along with it, and Delangue says that cost pressure is what eventually pushes them toward open source models they can run and adapt themselves. In his telling, this shift from renting to owning the underlying models is now a repeatable pattern rather than a one-off decision made by a handful of early adopters.

Delangue raised these dynamics in the context of Anthropic’s halted release of its Fable AI product, in a conversation with Rebecca Bellan on TechCrunch’s Equity podcast. He also raised a deeper concern: the possibility that a handful of large companies could end up controlling the entire AI landscape, concentrating power over the technology in very few hands. For Delangue, the growth of open models and datasets on platforms like Hugging Face works as a counterweight to that kind of concentration, spreading control over the building blocks of AI — the models and datasets themselves — across a wider set of companies rather than leaving it concentrated among a small number of frontier API providers. The Equity podcast episode is available on YouTube, Apple Podcasts, Overcast, Spotify and other platforms.

Why it matters

As companies scale their AI systems, the high costs of proprietary frontier APIs are driving them toward open source models, shifting the balance of power in the AI market.