Loading prices…
Lowest on-demand
$2.59per GPU·h · ≈ $1,891/month
at RunPodMedian on-demand
$2.92/GPU·h
across 5 providers
Lowest spot
$1.29/GPU·h
Reserved from $3.84
FP16/BF16 cost
$3.10/PFLOP·h
#47 of 70 GPUs for value
Providers
5
18 configurations
Price history
Daily H100 NVL prices per GPU-hour. The shaded period is reconstructed from the version history of SkyPilot's public price catalogs (fewer providers than live data); after that, PetaPrice's own tracking.
Lowest on-demandMedian on-demandLowest spot
Price by provider
Cheapest On-demand H100 NVL offer at each provider, per GPU-hour. Shorter and greener is cheaper.
RunPod🇺🇸$2.59best
GPU.ai🇺🇸$2.59best
Vast.ai🇺🇸$2.92+13%
Massed Compute🇺🇸$3.16+22%
Azure🇺🇸$6.982.7×
Thin vertical line = median across providers ($2.92).
Every provider we track
Pricing
Size
Host
Provider HQ
5 providers · 18 configurations
Sort
- 8× GPU · NVLink140 vCPU · 1,500 GB RAM · 10,000 GB disk$3.16 /GPU·h$25.3 /hOn-demand🇺🇸 US, Des Moines, IA○ sold outRent ↗
- 2× GPU · NVLink40 vCPU · 256 GB RAM · 2,500 GB disk$3.16 /GPU·h$6.32 /hOn-demand🇺🇸 US, Des Moines, IA○ sold outRent ↗
- 4× GPU · NVLink80 vCPU · 512 GB RAM · 5,000 GB disk$3.16 /GPU·h$12.6 /hOn-demand🇺🇸 US, Des Moines, IA○ sold outRent ↗
Alternatives to the H100 NVL
GPUs with similar compute and memory, and their lowest on-demand price right now.
H100 NVL pricing questions
How much does it cost to rent an H100 NVL?
On-demand H100 NVL prices start at $2.59 per GPU-hour (RunPod). The median across 5 providers is $2.92, and the most expensive is $6.98. Running one GPU around the clock costs about $1,891 per month at the lowest price.
Is there cheaper spot or interruptible H100 NVL capacity?
Yes. Spot (interruptible) H100 NVL capacity starts at $1.29 per GPU-hour, 50% below the cheapest on-demand price. Spot instances can be preempted at any time, so use them for checkpointed training or batch jobs.
How much compute do you get per dollar with the H100 NVL?
At $2.59 per hour and 0.835 PF of dense FP16/BF16, the H100 NVL costs $3.10 per PFLOP-hour. That ranks 47 of 70 GPUs with on-demand prices on PetaPrice. Figures are datasheet peaks; real utilization is typically 30–60%.
What are the H100 NVL's key specs?
94 GB HBM3 with 3,900 GB/s of memory bandwidth, Hopper architecture (2023), PCIe form factor, 400 W. Dense peak: 0.835 PF FP16 / BF16, 1.67 PF FP8, 1.67 PF INT8, 0.417 PF TF32, 0.06 PF FP32, 0.06 PF FP64.
How current are these prices?
Prices were collected on Sep 27, 2026, 01:12 UTC from provider APIs and public catalogs and refresh about every 2 hours. They are indicative list prices; always confirm on the provider's website before booking.