H100 SXM on CoreWeave
As of , CoreWeave lists the NVIDIA H100 SXM from $6.16 per GPU-hour on-demand, read . Among the 41 providers we track with an on-demand rate for this card, that is #36; the median rate card is $3.29, so CoreWeave is 87% above it. The cheapest is $1.79 on Cudo Compute. CoreWeave’s cheapest rate of any market is $2.44 (spot, Europe), capacity that can be taken back or shared, so it is not ranked as on-demand.
CoreWeave’s H100 SXM rates
On-demand
| GPU · what is sold | $ / GPU-hour | Read | Source |
|---|---|---|---|
| HGX H100$49.24/node ÷ 8 | $6.16 | captured | Deploy ↗ |
Spot / interruptible
| GPU · what is sold | $ / GPU-hour | Read | Source |
|---|---|---|---|
| HGX H100$19.51/node ÷ 8 · Europe · −$3.72 (−60%) vs its on-demand $6.16 | $2.44 | captured | Deploy ↗ |
| HGX H100$19.71/node ÷ 8 · North America · −$3.69 (−60%) vs its on-demand $6.16 | $2.46 | captured | Deploy ↗ |
Providers near this rate
Each provider’s cheapest live on-demand rate for the NVIDIA H100 SXM, CoreWeave among them, 6 of 41. The full list is on the NVIDIA H100 SXM page.
| Place | Provider | USD / GPU-hour | Read | Deploy |
|---|---|---|---|---|
| #1 |  Cudo Compute GPU only | $1.79 | captured | Deploy ↗ |
| #34 |  Tata Communications from | $5.16 | captured | Deploy ↗ |
| #35 | FluidCore | $5.55 | captured | Deploy ↗ |
| #36 | CoreWeave | $6.16 | captured | Deploy ↗ |
| #37 |  Amazon Web Services | $6.88 | captured | Deploy ↗ |
| #38 |  STACKIT | $8.99 | captured | Deploy ↗ |
More: every NVIDIA H100 SXM rate at every provider on the NVIDIA H100 SXM page (price history) · every card at CoreWeave on CoreWeave’s page · CoreWeave vs Cudo Compute, card by card. This page as data: /providers/coreweave/prices.json.
Data updated: