GPU.ai lists 5 GPU models across 5 configurations, starting at $0.520 per GPU-hour for the RTX 5090. It is the cheapest dedicated provider we track for 2 of its 5 GPU-and-pricing-model combinations: RTX 5090 (on-demand), A100 SXM (on-demand).
| GPU | VRAM | Pricing | GPU.ai $/GPU-hr | Best dedicated | Difference | Cheapest at | |
|---|---|---|---|---|---|---|---|
| RTX 5090 | 32 GB | on-demand | $0.520 | $0.520 | cheapest | — | rent |
| A100 SXM | 80 GB | on-demand | $1.07 | $1.07 | cheapest | — | rent |
| H100 SXM | 80 GB | on-demand | $3.99 | $1.79 | +123% | Cudo Compute | rent |
| H200 | 141 GB | on-demand | $5.29 | $2.91 | +82% | Seeweb | rent |
| B200 | 192 GB | on-demand | $7.99 | $4.09 | +95% | Together AI | rent |
How much does a GPU cost on GPU.ai?
GPU.ai starts at $0.520 per GPU-hour for the RTX 5090. It lists 5 GPU models in total: A100 SXM, B200, H100 SXM, H200, RTX 5090.
Is GPU.ai cheaper than other GPU clouds?
For 2 of the 5 GPU and pricing-model combinations it offers, GPU.ai is the cheapest dedicated provider on this table — including RTX 5090, A100 SXM. For the rest, the table above names the provider that beats it and by how much. Comparisons are like-for-like on GPU, pricing model and capacity tier.
Which GPUs does GPU.ai offer?
5 models: A100 SXM, B200, H100 SXM, H200, RTX 5090. Availability changes; the "in stock" flags on each GPU page reflect the last scrape.