Cheapest on-demand rate per card, 6 cards.
| GPU | On-demand / GPU-hour |
|---|---|
| NVIDIA L40S Everywhere Inference | $1.11 read |
| NVIDIA A100 80GB Bare Metal GPU Clusters | $1.51 read |
| NVIDIA H100 SXM Bare Metal GPU Clusters | $3.01 read |
| NVIDIA H100 80GB (form factor not stated) Virtual GPU | $3.12 read |
| NVIDIA H200 Bare Metal GPU Clusters | $3.52 read |
| NVIDIA B300 Bare Metal GPU Clusters | $6.63 read |