Crusoe Cloud GPU pricing 2026
Climate-aligned GPU cloud running on stranded gas and renewable power. Pure AI-infrastructure focus with H100, H200, and B200 SuperPODs.
Published rate card
Per-GPU per-hour rates pulled from the vendor pricing page linked above, Last verified June 2026. Rates exclude storage, egress, and any managed-service uplift. Reserved-capacity contracts typically improve on these rates.
| GPU | Configuration | Per GPU per hour |
|---|---|---|
| H100 SXM | On-demand, per-GPU on 8x | $4.650 |
| H200 SXM | On-demand | $6.500 |
| A100 80GB SXM | On-demand, per-GPU | $2.450 |
| L40S | On-demand | $1.500 |
| A40 | On-demand | $1.100 |
Hidden costs to watch
- Reserved capacity (3-12 month) unlocks lower rates; pure on-demand is positioned as the premium tier.
- Managed Kubernetes (CMK), object storage, and high-performance file system billed on top of GPU compute.
- Single-region availability matters; cross-region replication is the customer's design problem.
What Crusoe Cloud is best for
Sustainability-conscious training workloads that want stranded-energy data centres and H100 / H200 capacity.
See the GPU cloud buying guideWorked example
Acme Vision Co. (illustrative example, not a real company) needs to train a 10 billion-parameter vision-language model on a fixed 8-GPU cluster for 30 days at 18 hours per day duty cycle.
At Crusoe Cloud's cheapest published rate of $1.50 per GPU-hour, the run costs $6,480 for raw GPU compute, before storage, egress, and MLOps overhead. Add the typical 25 percent year-one uplift and the modelled spend is $8,100. Use the calculator on the homepage to model your own GPU class, cluster size, and duty cycle.
Last verified June 2026. Crusoe Cloud rates change frequently. Always obtain a vendor quote before purchase.
Visit Crusoe Cloud pricing page