GPU pricing calculator.

Compare GPU cloud costs across providers. Select your GPU, usage pattern, and see real-time cost estimates, then rent the GPU server.

Configure usage

GPU model

Your estimate

$252.00/mo

1 × RTX 4090 · 24h/day · 30 days · 720 GPU-hours

Hourly
$0.35/hr
Annual
$3,024/yr

You save $280.80/month−53%

53% less than RunPod, or $3,370/year at this usage.

Cost comparison across providers for 1 RTX 4090 GPU at 24 hours per day, 30 days per month
ProviderMonthlyvs OpenRelay
OpenRelay$252.00/mobaseline
RunPod$532.80/mo+$280.80/mo
AWS$871.20/mo+$619.20/mo
Lambda LabsN/AN/A

Per GPU-hour, no minimum

Stop anytime with no penalties or minimum commitments.

What the rate covers

Load balancing and the default disk are included in the hourly rate. Other disk storage is $0.096/GB-month. Data transfer through our gateway is $0.11/GB, and your VM's own internet traffic is free.

Failover included

Automatic failover to healthy nodes at no extra cost. Your workloads stay running even if nodes go down.

Ready to start saving on GPU compute?

Deploy any GPU in minutes with per GPU-hour pricing, automatic failover, and no minimum commitments.