H100 NVL pricing on Runpod
Every Runpod H100 NVL offering we track, normalized to a per-GPU hourly rate and ranked against 4 providers with current H100 NVL pricing.
Runpod is currently the cheapest H100 NVL among the 4 providers we track.
Runpod H100 NVL rates
Hourly prices per GPU. Multi-GPU rows show the per-GPU rate, not the instance total.
| Price / GPU-hr | Type | GPUs |
|---|---|---|
| $2.59 | On-demand | 1 |
| $2.59 | On-demand | 2 |
| $2.59 | On-demand | 3 |
| $3.07 | Spot | 1 |
| $3.07 | Spot | 2 |
| $3.07 | Spot | 3 |
| $3.07 | Spot | 4 |
| $3.07 | Spot | 5 |
| $3.07 | Spot | 6 |
| $3.07 | Spot | 7 |
| $3.07 | Spot | 8 |
| $3.19 | On-demand | 1 |
| $3.19 | On-demand | 2 |
| $3.19 | On-demand | 3 |
| $3.19 | On-demand | 4 |
| $3.19 | On-demand | 5 |
| $3.19 | On-demand | 6 |
| $3.19 | On-demand | 7 |
| $3.19 | On-demand | 8 |
H100 NVL specifications
- VRAM
- 94GB
- Architecture
- Hopper
- CUDA Cores
- 14,592
- Tensor Cores
- 456
- Memory Bandwidth
- 3958 GB/s
- TDP
- 400W
H100 NVL on other providers
About Runpod
Runpod is an AI developer cloud for building, deploying, and scaling AI workloads. It offers on‑demand GPU pods across 30+ regions in two tiers — data‑center‑hosted Secure Cloud and lower‑cost Community Cloud — plus serverless endpoints and multi‑node clusters, with B200 and H100/H200 alongside A100, L40S, and RTX classes.
Frequently Asked Questions
How much does the H100 NVL cost on Runpod?
Runpod lists the H100 NVL from $2.59 per GPU-hour across 19 tracked configurations. Prices are collected daily — see the table above for the current rates.
Is Runpod the cheapest place to rent the H100 NVL?
Of the 4 providers we track with current H100 NVL pricing, Runpod is currently the cheapest per GPU-hour. Rates change often, so check the comparison above before committing.
Which H100 NVL configurations does Runpod offer?
We currently track 19 H100 NVL offerings from Runpod, covering the GPU counts, pricing types, and regions listed in the table above. Every price is normalized to a per-GPU hourly rate so it can be compared across providers.