NVIDIA H100 NVL rental price
Frontier AITwo PCIe H100s bridged into a 94 GB pair, built for large-model inference in conventional servers.
live · updated 17d ago
VRAM
94 GB HBM3
Architecture
Hopper
Segment
Frontier AI
Providers quoting
4
best rate
$2.590/GPU-hr at RunPod
24h
—
7d
—
30d
—
cheapest 8× node
$25.28/hr
The cheapest NVIDIA H100 NVL you can rent right now is $2.590 per GPU-hour at RunPod, against $6.980 at the dearest of the 4 providers quoting it — a 2.7× spread for identical silicon. A full 8-GPU node starts at $25.28/hr at Massed Compute. Rates below refresh automatically.
Rent vs. buy
buy outright
$26,768
rent
$2.590/hr
breakeven
14.1 months
Buying a NVIDIA H100 NVL 94GB at today’s street price of $26,768 costs the same as 10,335 hours (14.1 months) of renting one at $2.590 an hour. Run it less than that and renting is simply cheaper; run it more and the card starts paying itself back — but this is a floor, not a total cost of ownership. It counts only the purchase price against the rental rate, and ignores power, cooling, rack space, networking, depreciation and the cost of tying up the capital, all of which push the real breakeven further out.
NVIDIA H100 NVL 94GB purchase prices and history →H100-NVL hourly rate history
Daily low $/GPU-hour per provider, cheapest provider first — hover the chart for detail.
Where to rent the NVIDIA H100 NVL
Vast.ai
Marketplace
$2.669/GPU-hr
1× H100-NVL · $2.669/hr per node · The Netherlands, NL
- 1× H100 NVL x1$2.670/hr
Massed Compute
GPU cloud
$3.160/GPU-hr
8× H100-NVL · $25.28/hr per node · pcie · US, Des Moines, IA
no capacity right now
- 2× gpu_2x_h100_nvl_nvlink · no capacity$6.320/hr
- 4× gpu_4x_h100_nvl_nvlink · no capacity$12.64/hr
Microsoft Azure
Hyperscaler
$6.980/GPU-hr
1× H100-NVL · $6.980/hr per node · eastus
- 2× NC80adis_H100_v5$13.96/hr
H100-NVL cluster nodes
Multi-GPU configurations, cheapest per GPU-hour first. For distributed training the interconnect matters as much as the rate — an NVLink or InfiniBand node and the same count of loose PCIe cards are not the same machine.
| Provider | Configuration | Interconnect | $/GPU-hr | Node/hr |
|---|---|---|---|---|
| Massed Computeno capacity | 8× gpu_8x_h100_nvl_nvlink · US, Des Moines, IA | pcie | $3.160 | $25.28/hr |
| Massed Computeno capacity | 2× gpu_2x_h100_nvl_nvlink · US, Des Moines, IA | pcie | $3.160 | $6.320/hr |
| Massed Computeno capacity | 4× gpu_4x_h100_nvl_nvlink · US, Des Moines, IA | pcie | $3.160 | $12.64/hr |
| Microsoft Azure | 2× NC80adis_H100_v5 · eastus | — | $6.980 | $13.96/hr |