Featherless.ai Inc., a serverless inference platform startup that hosts open-source artificial intelligence models, today revealed it has raised $20 million in new funding to expand its global infrastructure and launch a marketplace for specialized open models.

Founded in 2024, Featherless.ai pitches itself as a neutral hosting layer for enterprises that want to run open-source AI without committing to a single proprietary cloud or hardware stack. The company’s platform serves more than 30,000 open models spanning language, vision and audio and is billed as the fastest-growing inference partner on Hugging Face.

Customers access models through a single application programming interface, with Featherless.ai handling the underlying graphics processing unit allocation, scaling and routing. The company argues that its approach lowers the barrier for enterprises that want to test or deploy niche open models without standing up dedicated infrastructure for each one.

“We didn’t just build an inference engine, we built an AI optimization stack: inference, model and workflow optimization working together as a system,” the company said in a statement. “This is how we deliver performance and cost efficiency that closed platforms can’t match on a single model, let alone thirty thousand.”

Featherless.ai co-founder and Chief Executive Eugene Cheah is one of the creators of RWKV, an open-source model architecture developed under the Linux Foundation that uses a recurrent design as an alternative to transformer-based systems. He co-founded the company with Chief Technology Officer Harrison Vanderbyl and Chief Operating Officer Wesley George, with the founding team drawn from the RWKV open-source community.

The Series A round was co-led by AMD Ventures and Airbus Ventures, with BMW i Ventures, Kickstart Ventures, Panache Ventures and Wavemaker Ventures also participating.

Featherless.ai plans to use the new funding in three areas. It plans to scale infrastructure across additional regions, launch a dedicated marketplace for fine-tuned and specialized open models, and deepen integration with a wider range of hardware architectures, including AMD accelerators, to push inference costs lower.

The funding comes as enterprise buyers continue to weigh the cost and lock-in tradeoffs of frontier proprietary models from companies including OpenAI Group PBC and Anthropic PBC against a growing catalog of open-weights releases from Meta Platforms Inc., Mistral AI SAS, Alibaba Group Holding Ltd. and others.

Before today’s new funding, Featherless.ai had previously raised a $5 million seed round in 2025 with backing from Airbus Ventures.

Image: Featherless.ai