Independent reference. We are independent of every vendor listed. No affiliate links. No sponsored placements.
Vendor pricing - Last verified June 2026

Hyperstack GPU pricing 2026

Nexgen-Cloud Hyperstack is a European on-demand GPU cloud with per-minute billing and an aggressive H100 / A100 price point.

Pricing model
on-demand
Cheapest published rate
$0.50 per GPU-hour
RTX A6000
Public source
https://www.hyperstack.cloud/gpu-pricing
Last verified June 2026

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.

GPUConfigurationPer GPU per hour
H100 SXMOn-demand$2.400
H100 PCIeOn-demand$1.900
A100 80GB SXMOn-demand$1.350
A100 80GB PCIeOn-demand$1.350
L40On-demand$1.000
RTX A6000On-demand$0.500

Hidden costs to watch

  • Reserved contracts (1, 6, 12, 36 month) unlock the deepest discount tier.
  • Storage and networking metered separately; per-minute billing applies once an instance is running.
  • Capacity is sometimes constrained on the headline H100 SKUs during AI-training peak demand.

What Hyperstack is best for

Cost-sensitive H100 and A100 training where the European data-centre footprint and per-minute billing matter.

See the GPU cloud buying guide

Worked 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 Hyperstack's cheapest published rate of $0.50 per GPU-hour, the run costs $2,160 for raw GPU compute, before storage, egress, and MLOps overhead. Add the typical 25 percent year-one uplift and the modelled spend is $2,700. Use the calculator on the homepage to model your own GPU class, cluster size, and duty cycle.

Last verified June 2026. Hyperstack rates change frequently. Always obtain a vendor quote before purchase.

Visit Hyperstack pricing page