NVIDIA L40S pricing.
Prices checked
The cheapest L40S rents for $0.79/GPU-hr at RunPod, and the typical hour is $1.67 across 5 providers, checked .
Cheapest hour
RunPod
L40S · Community Cloud
The least you pay for one L40S for an hour on demand. Interruptible, it is $0.95/GPU-hr at Verda (DataCrunch).
Typical hour
the middle of 12 machines
Prices run from $0.79 to $2.25 for the same board; the rows below say what comes with each.
Every provider renting the L40S
- 1
L40S, Community Cloud · 16 vCPU
$0.79/GPU-hr - 2
L40S, Secure Cloud · 16 vCPU
$1.09/GPU-hr - 3
NVIDIA L40S
$1.50/GPU-hr - 4
vcg-l40s-16c-180g-48vram · 16 vCPU
$1.67/GPU-hr - 5
vcg-l40s-32c-375g-96vram · 2 GPUs, $3.34/hr all in
$1.67/GPU-hr - 6
vcg-l40s-64c-750g-192vram · 4 GPUs, $6.68/hr all in
$1.67/GPU-hr - 7
vbm-64c-2048gb-8-l40-gpu · 8 GPUs, $13.37/hr all in
$1.67/GPU-hr - 8
1L40S.20V · 20 vCPU
$1.90/GPU-hr - 9
2L40S.40V · 2 GPUs, $3.80/hr all in
$1.90/GPU-hr - 10
4L40S.80V · 4 GPUs, $7.60/hr all in
$1.90/GPU-hr - 11
8L40S.160V · 8 GPUs, $15.21/hr all in
$1.90/GPU-hr - 12
NVIDIA L40S · 8 GPUs, $18/hr all in
$2.25/GPU-hr
Not ranked
A serverless rate buys the hour a function runs, not a machine you hold: CPU and memory are billed on top, so the rate is a floor rather than a price. How we read these
Nvidia L40S
$1.95/GPU-hr
How we count
Every rate is for one GPU for one hour, so a node of eight sits beside a single card at a comparable number, with the whole machine's hourly price in the row.
The L40S's 48 GB of memory and its bandwidth are the maker's own figures, the same whoever rents it out. A provider's feed is trusted for its prices and its GPU counts, never for what the board is.
- Marketplace
- The provider matches you with someone else's machine.
- Own machines
- The provider owns or leases the machines.
- Serverless · not ranked
- Billed by the hour a function runs, not a machine you hold, so it is not ranked.
Similar GPUs
Other boards with about as much memory, cheapest hour first. Each percentage is against the L40S's cheapest hour, $0.79.
Frequently asked
- Who rents the L40S cheapest?
- RunPod, at $0.79/GPU-hr on L40S · Community Cloud, checked 1 Nov 2026. That is the on-demand rate for one GPU for one hour; the whole machine bills $0.79/hr.
- What is the NVIDIA L40S?
- 48 GB of GDDR6 at a fraction of an H100's price. Good for inference and rendering, slow for training. It is a Ada Lovelace board from 2023 with 48 GB of memory at 864 GB/s, in the PCIe card package. Those figures come from the maker, not from any provider's price page.
- Can I rent a L40S on spot?
- Yes. 5 of the 12 configurations here publish an interruptible rate, the cheapest being $0.95/GPU-hr at Verda (DataCrunch), 50% under its own on-demand rate. The machine can be taken back mid-run, so it suits work that checkpoints.
- Why do L40S prices differ so much?
- The spread here runs from $0.79/GPU-hr to $2.25/GPU-hr for the same silicon, because what else comes with it differs: the vCPU and RAM around the GPU, whether a disk is included, how many regions it runs in, and whether the machine belongs to the provider or to someone else on a marketplace. Each row's tag says whose it is.
- How current are these prices?
- They are read every 3 days from each provider's own public page or API; the readings on this page are from 1 Nov 2026. Reserved and committed rates are not listed. Prices exclude tax.
Same generation