Open source AI is booming. Hugging Face explains who profits.

TechCrunch's Equity podcast released an episode focused on the economics of open source AI, with host Rebecca Bellan sitting down with Irene Solaiman, Chief Policy Officer and SVP of Strategy at Hugging Face. The conversation is framed around a question the industry is only starting to confront: once capable models are widely and freely available, where does the money actually accrue?

The stated backdrop is adoption. Open source models are described as going mainstream, with enterprises, governments and startups alike drawn to the same three promises: lower costs, tighter control over deployments and greater flexibility in how models are shaped and hosted. That combination has made open weights a credible alternative to proprietary APIs for organizations that want to keep data, latency and customization decisions in-house.

But wider availability sharpens the commercial puzzle the episode sets out to examine. As increasingly capable models become free to download and run, companies must find ways to capture value around the model rather than on top of a scarce artifact — whether through hosting, tooling, evaluation, safety layers or services. The episode treats this as an open problem rather than a settled one.

A second thread is geopolitical. With Chinese open models gaining ground, the episode asks plainly whether American players can keep up — a question that sits at the intersection of model release cadence, compute access and policy, territory that overlaps directly with Solaiman's dual remit over strategy and policy at one of the largest open model hubs.

For builders, the practical agenda is explicit: what startups need to know before committing to open models, from selecting the right tools and infrastructure to navigating the trade-offs that will shape the future of open AI. The framing suggests the choice is less about model quality in isolation and more about the stack, maintenance burden and strategic dependencies that come with it.

What to watch is how these threads resolve in practice: which business models prove durable when the weights themselves are free, whether the gap between US and Chinese open releases widens or narrows, and how much infrastructure flexibility startups actually retain once they standardize on a particular open ecosystem.