News
DealsCorvex agrees USD 103 million take-or-pay deal to perform AI inference with Mid-Atlantic GPUsSovereignNeysa scales AI infrastructure in India with AMD Instinct MI350X GPUs on TytanInferenceMicrosoft Foundry adds Baseten, DeepSeek AI inference (MSFT:NASDAQ)ClustersDataOne and Rowan University sign engineering co-op program agreement for DataOne's Vineland AI factoryClustersLook inside Cal Poly's $3 million AI Factory (Video)DealsOpenAI Taps AMD Chips for 457 MW of AI Capacity in Six-Year TensorWave DealClustersCal Poly launches an AI data center on campus. Take a look insideInferenceArista takes aim at the next AI network battleground: distributed inferenceDealsOpenAI Signs 457 MW AI Data Center Deal as AMD Expands Its Challenge to NvidiaClustersBehind Dell AI Factory鈥檚 NVIDIA & AMD Enterprise ExpansionDealsCorvex agrees USD 103 million take-or-pay deal to perform AI inference with Mid-Atlantic GPUsSovereignNeysa scales AI infrastructure in India with AMD Instinct MI350X GPUs on TytanInferenceMicrosoft Foundry adds Baseten, DeepSeek AI inference (MSFT:NASDAQ)ClustersDataOne and Rowan University sign engineering co-op program agreement for DataOne's Vineland AI factoryClustersLook inside Cal Poly's $3 million AI Factory (Video)DealsOpenAI Taps AMD Chips for 457 MW of AI Capacity in Six-Year TensorWave DealClustersCal Poly launches an AI data center on campus. Take a look insideInferenceArista takes aim at the next AI network battleground: distributed inferenceDealsOpenAI Signs 457 MW AI Data Center Deal as AMD Expands Its Challenge to NvidiaClustersBehind Dell AI Factory鈥檚 NVIDIA & AMD Enterprise Expansion
EveryNeocloud
Featherless AI

Featherless AI

Serverless inference platform for open-source AI models

Featherless AI is a serverless inference platform that provides access to over 30,000 open-source models via a single API. Founded in 2023 by the creators of RWKV, it rebranded from Recursal.AI in 2024 and raised $20 million in 2026 to support vendor-neutral AI infrastructure.

Share on WhatsApp

  • Serves over 30,000 open-source models via one API
  • Raised $20M Series A in 2026 led by AMD and Airbus
  • Founded by co-leads of the RWKV Linux Foundation project
  • Operates in 19 countries with offices in SF, Toronto, Singapore

What it is

Featherless AI operates as a serverless inference platform designed to make open-source artificial intelligence models accessible through a single application programming interface. The company hosts more than 30,000 models, covering language, vision, and multimodal tasks. It positions itself as a neutral layer independent of major hyperscalers or proprietary ecosystems. Headquartered in San Francisco, the firm serves users across 19 countries and over 60 cities. Its primary offering allows developers to run models without managing underlying infrastructure, using flat monthly pricing.

How it came about

The organisation began in 2023 as a research team rather than an infrastructure provider. Its founders, Eugene Cheah, Harrison Vanderbyl, and Wesley George, previously co-led RWKV, a Linux Foundation project focused on attention-free model architectures. This research aimed to reduce the cost of running large-scale AI. The team initially operated under the name Recursal.AI. In 2024, the company rebranded to Featherless AI to launch its serverless inference platform. By 2025, it claimed to be the fastest-growing inference partner in the Hugging Face ecosystem. In 2026, it secured Series A funding to scale its operations globally.

The compute

Featherless provides serverless access to its model library, removing the need for users to manage servers or clusters. The platform supports tens of thousands of open-source models through one API. While specific hardware details are limited, the involvement of AMD Ventures suggests the use of AMD Instinct accelerators in its infrastructure. The system is built on research into efficient, attention-free architectures derived from the RWKV project. This technical foundation allows for cheaper and faster inference compared to traditional attention-based models. The company emphasises optimisation stacks developed by its founding team.

Where the GPUs are

The company maintains a remote-first operational model with its headquarters in San Francisco, California. It lists satellite offices in Toronto, Singapore, and various locations across Europe. Featherless states that it operates in 19 countries and serves users in more than 60 cities worldwide. This distributed approach supports its goal of providing global access to open-source AI infrastructure. The physical location of its compute resources is not explicitly detailed, but the service availability spans multiple regions to ensure low-latency access for international developers.

Customers and deals

Specific customer names are not disclosed in the available information. The platform targets developers and researchers who require easy access to open-source models. It identifies itself as a leading inference partner within the Hugging Face ecosystem. By serving more open-source models than other providers, it attracts users working with language, vision, and multimodal AI tasks. The value proposition focuses on simplicity and neutrality, appealing to those wishing to avoid vendor lock-in with major cloud providers.

Money and ownership

In 2026, Featherless AI raised a $20 million Series A round. The investment was co-led by AMD Ventures and Airbus Ventures. Other participants included BMW i Ventures, Kickstart Ventures, Panache Ventures, and Wavemaker Ventures. This capital is intended to scale the company鈥檚 open-source AI infrastructure worldwide. The funding supports the mission to keep AI infrastructure independent of any single vendor.

Its place in AI

Featherless plays a role in democratising access to open-source AI models. By leveraging research from the RWKV project, it contributes to the development of efficient, attention-free architectures. The company claims to host the largest model library in its ecosystem as of 2025. Its existence challenges the dominance of proprietary models by providing a neutral platform for community-created AI. This supports the broader open-source movement in artificial intelligence, ensuring that models from any creator remain accessible.

Questions people ask about Featherless AI

What is Featherless AI?

Featherless AI is a serverless inference platform that allows users to access over 30,000 open-source AI models through a single API. It aims to provide a neutral infrastructure layer independent of major hyperscalers.

Who founded Featherless AI?

The company was founded in 2023 by Eugene Cheah, Harrison Vanderbyl, and Wesley George. They previously co-led the RWKV project at the Linux Foundation, focusing on efficient AI architectures.

How much funding has Featherless raised?

In 2026, Featherless AI raised a $20 million Series A round. The investment was co-led by AMD Ventures and Airbus Ventures, with participation from several other venture firms.

Where is Featherless AI based?

The company is headquartered in San Francisco, California. It operates as a remote-first organisation with satellite offices in Toronto, Singapore, and across Europe.

Good reading and sources

    Where this page comes from

    There is no English Wikipedia article on Featherless AI yet, so the facts on this page come from the sources listed above: featherless.ai/about. Some are the subject's own pages, so its claims about itself are reported as its claims. EveryNeocloud wrote the text in its own words, with AI assistance, and compares it with the sources to make sure no passage is copied.

    It may contain mistakes, and figures change over time. Nothing here is investment or buying advice. If something is wrong, tell us. Last refreshed 2026-11-14.

    Photo credits on this page

    The small pictures above come from Wikimedia Commons under free licences. Some are cropped to fit.

    The weekly digest

    Every Monday: the provider of the day, new pages on EveryNeocloud, and the conferences and anniversaries of the week ahead.