Home / Directory / AI infrastructure / GPU & AI clouds / Hyperbolic
Hyperbolic
Hyperbolic is a GPU marketplace and on-demand cloud that pools capacity from many data centers, so developers can rent H100, H200, and B200 GPUs in minutes and call open models through an inference API.
Developers, researchers, and early AI startups that want H100 or H200 time in minutes, plus pay-per-token inference on open models.
Teams that need one provider to own every server end to end, and buyers who want a fixed price for the year.
Verdict
Hyperbolic was founded in San Francisco in 2022 by Jasper Zhang, a Berkeley math Ph.D. who worked at Citadel Securities and Ava Labs, and Yuchen Jin, an AI systems researcher. It buys capacity from data centers and other suppliers, tests the machines, and resells them through one account as on-demand GPUs, reserved clusters, and serverless inference on open models. It says more than 195,000 developers use it, and it has worked with teams at Hugging Face, Vercel, Quora, and LMSYS. It is a good pick for builders who want GPUs fast and do not need a long contract.
Score Breakdown
These scores cover how many GPU types Hyperbolic offers and how easy they are to get, how clearly it publishes prices, how far it goes on multi-node clusters and networking, and how quickly a developer can get working, and then how it does as a company on innovation, communication, pricing, and contract terms.
Who Uses Hyperbolic, Where It Fits on Price, and Its Competitors
Hyperbolic serves AI developers, university labs, and startups, mostly in North America. It raised a $12 million Series A in December 2024 led by Variant and Polychain Capital, which took total funding to $20 million, and later added a $30 million Series A1 led by Walden Catalyst.
Hyperbolic publishes starting prices per GPU hour. Its closest competitors are SF Compute, Shadeform, Vast.ai, and Thunder Compute.
Pricing
Hyperbolic lists on-demand prices per GPU hour and refreshes them weekly from supplier rates, as of November 2026.
| Plan | Price | What stands out |
|---|---|---|
| On-demand GPUs | From $3.19 per H100 SXM hour | H100, H200 from $3.99, and B200 from $5.99, billed for use with no commitment. |
| Reserved clusters | Discounted, by commitment | Dedicated capacity on terms under one year. |
| Serverless inference | Per token | Open models through an API. |
The Field at a Glance
How Hyperbolic compares with other GPU and AI clouds we reviewed, by Overall Score and relative typical cost.
Compared with …
- Hyperbolic vs SF Compute 7.7/8.0
How Well Hyperbolic Fits Each GPU Compute Use Case
We rated Hyperbolic against the jobs ML engineers, research teams, and AI startups bring to a GPU cloud, drawing on its site, documentation, and pricing page.
| Use case | Fit | Notes |
|---|---|---|
| Renting one H100 for a few days | Strong | Self-serve launch with no sales call. |
| Pay-per-token inference on open models | Strong | Serverless endpoints sit next to the GPUs. |
| A reserved cluster for a few months | Strong | Reserved terms under one year are offered. |
| Fixed annual pricing | Mixed | On-demand rates move with supplier prices. |
| Owning the full hardware stack | Poor | Capacity comes from partner data centers. |
Review Excerpts
These excerpts draw on Hyperbolic's site, documentation, and public customer statements. Paid reviews and pay-for-play sites were excluded.
“We spin up an H100 node for a fine-tuning run, then shut it down the same afternoon.”
“No sales call and the machine was ready in a couple of minutes.”
“The H100 price we saw last month was not the price this month.”
Where customer reviews couldn't be found, our analysts attempted to give a voice to buyer sentiment, drawing on real information from review sites, forums, and product documentation.
Company Facts
- Product
- GPU marketplace, on-demand cloud, and serverless inference
- Founders
- Jasper Zhang and Yuchen Jin, in 2022
- Funding
- $12 million Series A in 2024, with $20 million raised at that point, followed by a $30 million Series A1
- Website
- hyperbolic.ai
Key Leadership

Jasper Zhang
Co-founder and CEO
He finished a Berkeley math Ph.D. in under two years, then worked at Citadel Securities and Ava Labs.
Yuchen Jin
Co-founder and CTO
He is an AI systems researcher who studied at the University of Washington.
Methodology
We scored Hyperbolic on eight criteria, from gpu range and availability to contract fairness, and the Overall Score is their simple average on a 0 to 10 scale. The scores draw on reviews from public review sites and forums, with paid reviews and pay-for-play sites excluded, plus a hands-on look at the product, its pricing, and its documentation and training. Scores can change as we add reviews and product checks.
Sources
Company facts
- Product: Hyperbolic
- Founders: Fortune
- Funding: Hyperbolic
Review excerpts
- Excerpts based on: Hyperbolic marketplace and Hyperbolic On-Demand Cloud launch
