Hugging Face raised $235 million at a $4.5 billion valuation, according to TechCrunch, in a round that included Google, Amazon, Nvidia and Microsoft as investors. The financing confirmed the company's role as the default open repository for machine-learning models.
The position is strategically odd and commercially powerful. Hugging Face is the place the open-source AI community shares models, and it is funded by the closed-platform companies whose models that community often competes with.
Why the structure works
Every hyperscaler has a commercial interest in the open ecosystem existing, because the open ecosystem trains the engineers who eventually buy cloud capacity. Funding the canonical repository is cheaper than building it, and it keeps the ecosystem on infrastructure the hyperscalers can sell into.
The commercial question
The company has to convert its position as the default sharing layer into a paid product, which means enterprise hosting, private model registries and inference. The early data suggests that conversion is working, and that the open ecosystem is a more reliable funnel than the closed one the hyperscalers are simultaneously trying to build.

