Featherless.ai secures $20M Series A to scale serverless hosting for open-source AI models

2 Sources

Share

Featherless.ai announced $20 million in Series A funding led by AMD Ventures and Airbus Ventures to expand its serverless inference platform. The startup hosts over 30,000 open-source AI models and plans to build a marketplace for specialized models while reducing infrastructure costs for enterprises seeking alternatives to proprietary AI systems.

Featherless.ai Secures Series A Funding to Expand Open-Source AI Infrastructure

Featherless.ai has raised $20 million in Series A funding led by AMD Ventures and Airbus Ventures, marking a significant milestone for the serverless inference platform startup

1

. The funding round, announced on April 30, 2026, also drew participation from BMW i Ventures, Kickstart Ventures, Panache Ventures, and Wavemaker Ventures

2

. Founded in 2024, the company positions itself as a neutral infrastructure for open-source AI that enables enterprises to deploy models without locking into proprietary systems or specific hardware stacks

1

.

Source: SiliconANGLE

Source: SiliconANGLE

Building a Cost-Efficient Hosting Layer for Thousands of Models

The startup's platform currently serves more than 30,000 open models spanning language, vision, and audio capabilities, establishing itself as the fastest-growing inference partner on Hugging Face

1

. Customers access these models through a single application programming interface, while Featherless.ai manages the underlying graphics processing unit allocation, scaling, and routing behind the scenes. This serverless hosting for open-source AI models eliminates the need for enterprises to build dedicated infrastructure for each model they want to test or deploy

1

.

"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 stated, emphasizing how this approach delivers performance and cost efficiency that closed platforms struggle to match across such a vast model catalog

1

.

Leadership Team Rooted in Open-Source Innovation

Co-founder and Chief Executive Officer Eugene Cheah brings deep expertise from creating RWKV, an open-source model architecture developed under the Linux Foundation that uses a recurrent design as an alternative to transformer-based systems

1

. He co-founded Featherless.ai with Chief Technology Officer Harrison Vanderbyl and Chief Operating Officer Wesley George, with the founding team emerging from the RWKV open-source community

1

.

"When a few dominant players control the entire stack, it stifles competition and limits what developers can imagine," Cheah explained. "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"

2

.

Strategic Plans to Reduce Inference Costs and Expand Access

Featherless.ai plans to deploy the new capital across three strategic priorities. The company will scale infrastructure across additional regions to improve global access to state-of-the-art models, launch a dedicated marketplace for specialized open models that have been fine-tuned for specific use cases, and deepen integration with a wider range of hardware architectures, including AMD accelerators, to reduce inference costs further

1

.

The marketplace for specialized open models represents a particularly notable development, as it could enable enterprises to discover and deploy domain-specific AI capabilities without building custom infrastructure. By supporting diverse hardware architectures beyond traditional options, the platform aims to drive down costs while maintaining performance across different deployment scenarios

1

.

Timing Aligns with Growing Enterprise Demand for Alternatives to Proprietary Systems

The funding arrives as enterprise buyers increasingly evaluate tradeoffs between frontier proprietary models from companies like OpenAI and Anthropic versus a growing catalog of open-weights releases from Meta Platforms, Mistral AI SAS, Alibaba Group Holding, and others

1

. Organizations are weighing cost considerations and vendor lock-in risks against the performance and support offered by closed systems.

By ensuring that state-of-the-art models remain accessible outside proprietary "walled gardens," Featherless.ai provides developers with the creative flexibility to build the next generation of applications

2

. This serverless approach lowers the barrier for enterprises wanting to experiment with niche open models without committing significant resources upfront

1

.

Before this Series A round, Featherless.ai had previously raised a $5 million seed round in 2025 with backing from Airbus Ventures

1

. The continued support from Airbus Ventures, now joined by AMD Ventures as co-lead, signals strong investor confidence in the open-source AI infrastructure market and the company's execution capability.

Today's Top Stories

TheOutpost.ai

Don’t drown in AI news. We cut through the noise - filtering, ranking and summarizing the most important AI news, breakthroughs and research daily. Spend less time searching for the latest in AI and get straight to action.

Instagram logo
LinkedIn logo
Youtube logo
Ā© 2026 TheOutpost.AI All rights reserved