Hyperbolic, a startup focused on decentralized AI computing and inference services, announced on Tuesday it raised $12 million in a Series A funding round led by Variant and Polychain Capital. 

The round drew participation from a wide array of investors, including Lightspeed Faction, Bankless Ventures, and Wintermute Ventures. Angel investors included Sreeram Kannan of EigenLayer and Alex Atallah of OpenRouter.

It brings the company’s total funding to $20 million as it seeks to expand its footprint in the burgeoning AI and Web3 markets, the company said in a statement.

The San Francisco-based firm is building a decentralized GPU marketplace and offering verifiable AI inference services, addressing what it sees as a critical shortage of accessible and affordable computing power for developers and researchers.

Hyperbolic’s GPU marketplace allows suppliers to monetize idle GPUs while providing renters with cost savings of up to 75%. The platform is powered by Hyper-dOS, its decentralized operating system. 

Its inference service, built on the proprietary Proof of Sampling (PoSP) protocol, touts the ability to process over 1 billion tokens daily for 30,000 Web2 users, the company said.

Founded by Jasper Zhang, Ph.D., and Yuchen Jin, Ph.D., both AI and mathematics researchers, Hyperbolic aims to bridge Web2 and Web3 with solutions that unify fragmented AI workflows. 

“Hyperbolic is the first player we've encountered that truly addresses the 'cost of trust' issue in decentralized GPU networks without sacrificing performance, quality, and user experience,” Jesse Walden, managing partner at Variant, said. “With Jasper and Yuchen’s deep expertise spanning both crypto and AI, they are uniquely equipped to transform decentralized compute marketplaces for AI models.”

The company plans to launch its own blockchain in 2025 to facilitate collective governance for AI infrastructure.

Hyperbolic’s services are already utilized by major AI players, including Hugging Face, Quora, and leading universities such as Stanford and NYU.

The company said it will use the funds to grow its engineering and go-to-market teams and scale its decentralized AI platform globally.