H200 pricing on Crusoe
Every Crusoe H200 offering we track, normalized to a per-GPU hourly rate and ranked against 31 providers with current H200 pricing.
The cheapest H200 we currently track is $2.00/hr on Verda.
Crusoe H200 rates
Hourly prices per GPU. Multi-GPU rows show the per-GPU rate, not the instance total.
| Price / GPU-hr | Type | GPUs |
|---|---|---|
| $4.29 | On-demand | 1 |
H200 specifications
- VRAM
- 141GB
- Architecture
- Hopper
- CUDA Cores
- 16,896
- Tensor Cores
- 528
- Memory Bandwidth
- 4800 GB/s
- TDP
- 700W
H200 on other providers
- Verda$2.00/hr
- fal.ai$2.10/hr
- Nebius$2.45/hr
- CoreWeave$2.58/hr
- GMI Cloud$2.60/hr
- Deep Infra$2.69/hr
- Genesis Cloud$2.80/hr
- Civo$2.99/hr
- Koyeb$3.00/hr
- Seeweb$3.00/hr
- IO.NET$3.36/hr
- Massed Compute$3.44/hr
- Atlas Cloud$3.50/hr
- Sesterce$3.52/hr
- Runpod$3.59/hr
- Theta EdgeCloud$3.69/hr
- Denvr Dataworks$3.87/hr
- Hyperbolic$3.99/hr
- Hyperstack$3.99/hr
- Jarvis Labs$3.99/hr
- Lyceum$3.99/hr
- Oblivus$3.99/hr
- AceCloud$4.02/hr
- DigitalOcean$4.47/hr
- Modal$4.54/hr
- Together AI$6.60/hr
- Amazon AWS$7.91/hr
- AtmosCompute$9.84/hr
- Oracle Cloud$10.00/hr
- Microsoft Azure$10.60/hr
About Crusoe
Crusoe Cloud offers energy-first, high-performance GPU compute (H100, H200, A100, L40S, B200, GB200, and AMD MI300X/MI355X) alongside managed inference and fine-tuning services, backed by vertically integrated AI data centers.
Frequently Asked Questions
How much does the H200 cost on Crusoe?
Crusoe lists the H200 from $4.29 per GPU-hour across 1 tracked configuration. Prices are collected daily — see the table above for the current rates.
Is Crusoe the cheapest place to rent the H200?
Crusoe ranks #24 of 31 providers we track with current H200 pricing. The cheapest right now is Verda. Price is only one factor — region availability, network, and storage costs differ between providers.
Which H200 configurations does Crusoe offer?
We currently track 1 H200 offering from Crusoe, covering the GPU counts, pricing types, and regions listed in the table above. Every price is normalized to a per-GPU hourly rate so it can be compared across providers.