Transparent Pricing
Simple, predictable pricing for GPU infrastructure. No hidden fees. Volume discounts available.
On-Demand
Flexible hourly billing with no commitment. Spin up and down as needed.
- No minimum commitment
- Pay per hour of usage
- All GPU types available
- Standard support included
Monthly
Dedicated servers on a monthly basis. Lower rates with predictable billing.
- Month-to-month terms
- Dedicated hardware
- Priority support
- Volume discounts available
Reserved
Best pricing with annual commitments. Ideal for steady-state workloads.
- Up to 23% savings
- Guaranteed capacity
- Dedicated account manager
- Custom SLAs available
GPU Pricing
Per GPU pricing across commitment levels. All prices in USD.
| GPU Model | On-Demand | Monthly | Reserved (1yr) | Savings |
|---|---|---|---|---|
| NVIDIA H200 | $4.99/hr | $2,699/mo | $2,099/mo | 22% |
| NVIDIA H100 | $3.49/hr | $1,899/mo | $1,499/mo | 21% |
| NVIDIA A100 80GB | $1.99/hr | $1,099/mo | $849/mo | 23% |
| NVIDIA L40S | $0.99/hr | $549/mo | $429/mo | 22% |
Need higher volume? Contact us for custom pricing.
Frequently Asked Questions
How does billing work?
On-demand servers are billed hourly. Monthly and reserved commitments are billed at the start of each billing period. All usage is metered and invoiced accordingly.
What is included with the server?
All servers include full root access, dedicated networking, IPMI/BMC access, and standard technical support. Enterprise support and custom SLAs are available on reserved plans.
Can I scale up or down?
On-demand servers can be provisioned and decommissioned as needed. Monthly plans can be adjusted at the start of each billing cycle. Reserved plans offer guaranteed capacity.
Do you offer multi-GPU and multi-node configurations?
Yes. We offer 8-GPU nodes connected via NVLink and multi-node clusters connected via InfiniBand. Contact our sales team for custom cluster configurations.