
Two GPU specialist providers, priced against each other on the 11 GPUs both of them rent. Published per-GPU on-demand rates as each provider last stated them, with what their own pages say about billing, commitment and egress beside the numbers.
GPU.ai is cheaper on 8 of the 11 cards both of them rent, by a median of 18%. The two match on 3 of the rest.
The 11 cards both of them rent, cheapest published per-GPU on-demand rate each. Dearest first.
| GPU | GPU.ai | RunPod Secure Cloud | Difference |
|---|---|---|---|
| NVIDIA H200 SXM5 141GB | $3.59/hr | $4.59/hr | 22% cheaper on GPU.ai |
| NVIDIA H100 SXM5 80GB | $3.49/hr | $3.49/hr | same |
| NVIDIA H100 NVL 94GB | $3.19/hr | $3.19/hr | same |
| NVIDIA H100 PCIe 80GB | $2.48/hr | $2.89/hr | 14% cheaper on GPU.ai |
| NVIDIA RTX PRO 6000 Blackwell 96GB | $1.99/hr | $2.09/hr | 5% cheaper on GPU.ai |
| NVIDIA L40S 48GB | $0.80/hr | $1.09/hr | 27% cheaper on GPU.ai |
| NVIDIA L40 48GB | $0.78/hr | $0.82/hr | 5% cheaper on GPU.ai |
| NVIDIA RTX 6000 Ada 48GB | $0.65/hr | $0.84/hr | 23% cheaper on GPU.ai |
| NVIDIA A40 48GB | $0.49/hr | $0.49/hr | same |
| NVIDIA RTX A6000 48GB | $0.40/hr | $0.53/hr | 25% cheaper on GPU.ai |
| NVIDIA L4 24GB | $0.44/hr | $0.49/hr | 10% cheaper on GPU.ai |
Per GPU per hour, on-demand list prices as each provider last published them. Rates exclude tax and vary by region; a cheaper rate is not the same machine, the same network or the same support.
| GPU.ai | RunPod Secure Cloud | |
|---|---|---|
| Billing granularityOn short jobs this outweighs the rate: an hour minimum on a 90-second run costs more than a 30% dearer per-second rate. | Per second | Per second |
| Minimum commitment | $5 minimum deposit, no term | None |
| Payment model | Prepaid credit balance with burn rate and runway | Prepaid credits; post-paid invoice for Enterprise |
| Free credits | not stated | $1,000 startup credits (application); referral rewards |
| Egress | not stated | Free |
| Uptime SLA | not stated | 99.99% SLA on Enterprise agreements; none on Community Cloud |
| Regions | 8 regions: US East/Central/West, Canada, EU West, EU East, India South, Asia-Pacific | 31 regions across the US, Europe, Asia and Australia |
Neither site tells you whether your data trains their models, whether you get credits when it breaks, how fast support answers or how long you have to dispute a bill. We looked on both and record the absence rather than filling it in — if you need any of it, ask before you commit.
GPU.ai is cheaper on 8 of the 11 cards both of them rent, by a median of 18%. The two match on 3 of the rest.
NVIDIA H200 SXM5 141GB, NVIDIA H100 SXM5 80GB, NVIDIA H100 NVL 94GB, NVIDIA H100 PCIe 80GB, NVIDIA RTX PRO 6000 Blackwell 96GB, NVIDIA L40S 48GB, NVIDIA L40 48GB, NVIDIA RTX 6000 Ada 48GB, NVIDIA A40 48GB, NVIDIA RTX A6000 48GB, NVIDIA L4 24GB — 11 in total, on published per-GPU on-demand rates.
Of the cards both rent, GPU.ai starts at $0.40 per GPU-hour and RunPod Secure Cloud at $0.49.