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.

7.7/10
Overall Score
Recommended

Hyperbolic makes it quick to rent a GPU without a sales call and adds cheap open-model inference, though its prices move weekly with supplier rates and it does not own the hardware.

Best for

Developers, researchers, and early AI startups that want H100 or H200 time in minutes, plus pay-per-token inference on open models.

Not ideal for

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.

Buyer outcomes

GPU range and availability
7.8
Price transparency
7.4
Cluster and networking depth
7.5
Developer experience
8.2

Company & commercial

Innovation & product leadership
8.1
Project management & communication
7.4
Pricing
7.9
Contract fairness
7.3

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.

PlanPriceWhat stands out
On-demand GPUsFrom $3.19 per H100 SXM hourH100, H200 from $3.99, and B200 from $5.99, billed for use with no commitment.
Reserved clustersDiscounted, by commitmentDedicated capacity on terms under one year.
Serverless inferencePer tokenOpen 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.

6 7 8 9 10 Overall Score $ $$ $$$ $$$$ Relative typical engagement cost Fluidstack Nebius Shadeform CoreWeave CUDO Compute Verda Denvr Massed Compute Hot Aisle SF Compute Thunder Compute Sesterce Salad Hydra Host Hyperbolic 7.7
Hyperbolic Nebius CoreWeave Shadeform Verda SF Compute Thunder Compute Fluidstack Massed Compute Salad Hot Aisle Hydra Host Denvr Sesterce CUDO Compute

Compared with …

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 caseFitNotes
Renting one H100 for a few daysStrongSelf-serve launch with no sales call.
Pay-per-token inference on open modelsStrongServerless endpoints sit next to the GPUs.
A reserved cluster for a few monthsStrongReserved terms under one year are offered.
Fixed annual pricingMixedOn-demand rates move with supplier prices.
Owning the full hardware stackPoorCapacity 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.

How it's used

“We spin up an H100 node for a fine-tuning run, then shut it down the same afternoon.”

Representative buyer view
What people like

“No sales call and the machine was ready in a couple of minutes.”

Representative buyer view
What people don't like

“The H100 price we saw last month was not the price this month.”

Representative buyer view

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

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.

← Back to GPU & AI clouds

Sources

Company facts

Review excerpts