Comparison · GPU vs GPU · read 00:32 UTCH200 NVL vs GH200
Specifications, today's prices from 17 providers, price per unit of memory and compute, and who rents each cheapest. Better values are highlighted; lower is better for prices.
Side by sideSpecs, prices, value
Highlighted cells are the better of the two: more memory and compute, lower prices, lower cost per unit.
| H200 NVL | GH200 | Difference | |
|---|---|---|---|
| Memory & compute | |||
| VRAM | 141 GB | 96 GB | +47% |
| Memory bandwidth | 4.80 TB/s | 4.00 TB/s | +20% |
| FP16 tensor | 835.5 TFLOPS | 989 TFLOPS | -16% |
| FP32 | 60 TFLOPS | 67 TFLOPS | -10% |
| Power | 600 W | 1000 W | -40% |
| Prices today | |||
| Cheapest on-demand | $2.67 | $2.29 | +16% |
| Median on-demand | $4.14 | $2.75 | +50% |
| Highest on-demand | $8.88 | $6.50 | +37% |
| Cheapest spot | $1.33 | — | — |
| Cheapest reserved | $2.38 | — | — |
| Current offers (n) | 27 | 5 | +440% |
| Providers with a fresh offer | 12 | 5 | +140% |
| Value | |||
| Per GB of VRAM · hour (cheapest) | $0.019 | $0.024 | -21% |
| Per GB of VRAM · hour (median) | $0.029 | $0.029 | +2% |
| Per PFLOP FP16 · hour (median) | $4.949 | $2.779 | +78% |
| Per TB/s · hour (median) | $0.861 | $0.687 | +25% |
| Trend | |||
| 30-day median change | +27.2% | — | — |
| 90-day median change | +27.2% | — | — |
Cheapest providers today · on-demand · current offersWho rents each one cheapest
- 1
Vast.aiTrust 784× $2.67 - 2
Airon AI1× $2.70 - 3
IONOS2× $3.26
- 1
LambdaTrust 801× $2.29 - 2
Sesterce CloudTrust 601× $2.52
- 3
SpheronTrust 531× $2.75
Verdict · computed from the numbers aboveChoose one, choose the other
Choose the H200 NVL if…
- your model needs more than 96 GB per GPU: 141 GB against 96 GB.
- you want choice: 12 providers have a fresh offer today against 5.
Choose the GH200 if…
- the hourly price matters most: median $2.75 against $4.14, 34% less.
- you serve LLMs: memory bandwidth per dollar is better ($0.687 per TB/s-hour against $0.861).
Questions about H200 NVL vs GH200
Which is cheaper to rent, the H200 NVL or the GH200?
Today the cheapest current on-demand H200 NVL is $2.67 per GPU-hour and the cheapest GH200 is $2.29; medians are $4.14 and $2.75. Cheapest excludes offers marked unavailable and counts one floor per marketplace.
How much memory do the H200 NVL and GH200 have?
141 GB of HBM3e on the H200 NVL, 96 GB of HBM3 on the GH200, at 4.80 and 4.00 TB/s.
Where do these numbers come from?
From 32 current offers read on the providers' own pages and APIs over the last 45 days (reference 00:32 UTC), or in our October 2026 market study for the few providers our collector does not read. Specifications come from manufacturer datasheets. Every price on the GPU pages links to its receipt and shows its read date.