NVIDIA L40 rental price
DatacenterThe L40S's graphics-oriented sibling — same 48 GB, less tensor throughput, cheaper by the hour.
live · updated 17d ago
VRAM
48 GB GDDR6
Architecture
Ada Lovelace
Segment
Datacenter
Providers quoting
3
best rate
$0.335/GPU-hr at Vast.ai
24h
—
7d
—
30d
—
cheapest 8× node
$8.000/hr
The cheapest NVIDIA L40 you can rent right now is $0.335 per GPU-hour at Vast.ai, against $1.000 at the dearest of the 3 providers quoting it — a 3.0× spread for identical silicon. A full 8-GPU node starts at $8.000/hr at Hyperstack. Rates below refresh automatically.
L40 hourly rate history
Daily low $/GPU-hour per provider, cheapest provider first — hover the chart for detail.
Where to rent the NVIDIA L40
Vast.ai
Marketplace
$0.335/GPU-hr
2× L40 · $0.670/hr per node · Vietnam, VN
- 1× L40 x1$0.337/hr
Massed Compute
GPU cloud
$0.880/GPU-hr
1× L40 · $0.880/hr per node · pcie · US, Des Moines, IA
- 2× gpu_2x_l40$1.980/hr
- 4× gpu_4x_l40$5.960/hr
Hyperstack
GPU cloud
$1.000/GPU-hr
8× L40 · $8.000/hr per node · pcie · CA, Montreal
- 1× n3-L40x1$1.000/hr
- 2× n3-L40x2$2.000/hr
- 4× n3-L40x4$4.000/hr
L40 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 |
|---|---|---|---|---|
| Vast.ai | 2× L40 x2 · Vietnam, VN | — | $0.335 | $0.670/hr |
| Massed Compute | 2× gpu_2x_l40 · US, Des Moines, IA | pcie | $0.990 | $1.980/hr |
| Hyperstack | 8× n3-L40x8 · CA, Montreal | pcie | $1.000 | $8.000/hr |
| Hyperstack | 2× n3-L40x2 · CA, Montreal | pcie | $1.000 | $2.000/hr |
| Hyperstack | 4× n3-L40x4 · CA, Montreal | pcie | $1.000 | $4.000/hr |
| Massed Compute | 4× gpu_4x_l40 · US, Des Moines, IA | pcie | $1.490 | $5.960/hr |