Featherless.ai raises USD $20 million to expand open AI

Featherless.ai has raised USD $20 million in a Series A funding round co-led by AMD Ventures and Airbus Ventures.

BMW i Ventures, Kickstart Ventures, Panache Ventures and Wavemaker Ventures also participated as Featherless.ai looks to expand infrastructure for running open-source AI models.

The company will use the funding to grow its global infrastructure, build a marketplace for specialised open models and deepen integration with a range of hardware architectures. It positions itself as a neutral layer for AI deployment, untied to any single cloud provider, chipmaker or proprietary software environment.

The San Francisco-based group runs a serverless inference platform that supports more than 30,000 open models across language, vision and audio. It is also described as the fastest-growing Hugging Face inference partner.

Open model focus

Featherless.ai is targeting businesses that want to use open-source AI models while retaining more control over where systems run and where data is held. Its infrastructure is hosted in the US and Europe, and its team is spread across Canada, Europe, the US, Singapore and Australia.

That geographic footprint reflects a broader push among AI suppliers and customers to address concerns over jurisdiction, data privacy and the concentration of critical computing infrastructure among a small number of providers. Featherless.ai argues that open models and a wider mix of hardware offer companies an alternative path.

Hardware diversity is central to that strategy. Through a collaboration with AMD, popular open-source models run natively on the AMD ROCm software platform, offering an option beyond proprietary hardware systems.

The founding team also has roots in AI research. Its founders created RWKV, an open-source model architecture designed to challenge the long-standing dominance of transformer-based systems.

For investors, the appeal appears to lie in the prospect that more businesses will want a broader choice of models and infrastructure as AI deployments move from experimentation into routine use. The funding comes as debate continues across the sector over cost, access to computing resources and the influence of a small group of dominant technology companies.

"When a few dominant players control the entire stack, it stifles competition and limits what developers can imagine. We're building the infrastructure that makes open-source AI practical and reliable at scale, ensuring that enterprises can build on a foundation they actually own rather than one they merely rent."

This investment signals a turning point in the AI market. While the first wave of adoption was defined by proprietary, closed-door ecosystems, we provide neutral ground for a second phase in which companies can own and run their own models without being tethered to a single cloud provider or a restricted tech stack," said Eugene Cheah, Chief Executive and Co-founder of Featherless.ai.

Investor backing

Backers of the round pointed to rising demand from companies for greater control over cost, performance and model choice.

"Featherless.ai is at the forefront of a critical new phase in the development of the AI industry. By providing a strong foundation for open-source AI, it helps expand access and supports a more competitive and diverse ecosystem. We are delighted to support Featherless.ai on their journey," said Paz.

BMW i Ventures also tied the investment to enterprise demand for flexibility in deployment.

"As AI adoption accelerates, enterprises want more control over performance, cost and where their data lives. Featherless.ai is making leading open models production-ready at scale. Being able to use a variety of different models is key for future enterprise AI use cases. Featherless.ai is a key enabler in making this a reality," said Sage.

Airbus Ventures highlighted the economics of serving specialised models rather than relying solely on a small number of general-purpose systems.

"The next wave of AI adoption will not be carried by a handful of general-purpose models, but by the millions of specialized, fine-tuned models already being built in the open. The challenge has never been whether those models exist; it's whether they can be served reliably and at a cost that makes production deployment viable. Featherless.ai has solved the underlying architectural problem-loading models in seconds while keeping GPUs productively utilized-which is what finally makes the long tail commercially feasible. Our conviction in the team has only deepened since we first backed them, and we're proud to help lead this next phase," said Hitchcock.

Kickstart Ventures, an existing investor, said Featherless.ai's focus on inference infrastructure remains central to its thesis.

"Featherless.ai embodies what we believe about AI-it's infrastructure, not a vertical. We backed them at seed because of their technical edge in inference, and this round reinforces our conviction that they're building something durable in an increasingly competitive space," said Yao.

Panache Ventures cast the company as part of a broader effort to avoid excessive concentration in the AI market.

"Featherless.ai is one of those very rare startups that combines incredibly innovative and practical technology with an important and worthy company mission. The risk of the AI industry ultimately becoming dominated by a few players from one country is very real-solutions such as Featherless.ai provide a crucial alternative pathway for development," said Matta.