Performances
RD Popularity Index
RD Growth Index
Comparison
Follow Updates
Follow Lists
People
About Hyperbolic Labs
Hyperbolic is an AI cloud for developers that aggregates GPUs from multiple suppliers into a pool that can be provisioned on demand, offering hourly H100, H200 and B200 instances, OpenAI-compatible serverless inference and reserved clusters. It targets the quota approvals and long commitments of capacity buying: instances are reachable over SSH within minutes, billed hourly, with rates reset weekly against supplier pricing, while inference is priced per token across more than 25 open models.
Verification runs on Proof of Sampling and spML, which audit inference outputs by sampling and replace per-request zkML proofs with game-theoretic incentives to hold overhead down; Hyper-dOS handles cross-region scheduling, self-healing and image management. Customers are AI startups, enterprise inference teams and academic groups. Forge, an infrastructure layer that manages machine lifecycles across dozens of suppliers, went live in June 2026. Capacity depends on third-party supplier inventory and quotes, so available regions and prices widen or contract with them.
Founded in late 2022 by Yuchen and Jasper, Hyperbolic Labs began with a $725K pre-seed round. In July 2024, it raised $7M in seed funding led by Polychain Capital and Lightspeed Faction to build a decentralized GPU marketplace. December 2024 brought a $12M Series A, totaling $20M in funding. In February 2025, Hyperbolic became an official inference provider on Hugging Face, marking a major distribution milestone. By March 2025, it supported DeepSeek-V3-0324 on Hugging Face and saw 2.6x monthly GPU hour growth. In April 2025, the platform surpassed 195,000 developers processing 112 billion tokens monthly, and launched Auto Top-Ups and custom port configuration. May 2025 saw the release of an MCP server for natural language GPU orchestration and a rebrand to hyperbolic.ai.
In 2026 Hyperbolic closed a 12 million dollar Series A led jointly by Variant and Polychain Capital, with Chapter One, Lightspeed Faction, IOSG, Vertex, GSR and Wintermute Ventures among the participants, taking total funding to 20 million dollars. The capital goes to engineering headcount and to building its own chain, extending the compute marketplace from matching into settlement.
On technology, Hyperbolic validates inference through a sampling-based proof combined with economic incentives, arguing the approach beats zero-knowledge and optimistic machine learning on cost. Its Hyper-dOS orchestration layer schedules global GPU capacity with automatic scaling and self-healing. Products cover on-demand GPU rental, inference APIs and model hosting for research groups and cost-sensitive startups. The team numbers about 35 people and is based in New York.
Future strategy and roadmap: Hyperbolic plans to pull GPU supply, inference verification and payment settlement into one network that third-party providers can join, while expanding enterprise accounts and hosted open models so verifiable inference becomes the differentiator against centralised clouds.
Jasper Zhang — Co-founder and CEO; PhD in mathematics from UC Berkeley completed in two years; gold medalist at the Alibaba Global Math Competition and the Chinese Mathematical Olympiad; previously at Ava Labs and Citadel Securities.
Yuchen Jin — Co-founder and CTO; PhD in computer systems and networking from the University of Washington; led Relax, the open-source AI compiler within Apache TVM; published at NeurIPS, ICML and CVPR.
Hyperbolic Labs has seen recent team changes. Co-Founder & CTO Yuchen Jin left in April 2026, and Head of Product Kai Huang departed in May 2026.