NVIDIA L40S rental price per hour
48GB GDDR6 VRAM · Ada Lovelace architecture · released 2023
The cheapest way to rent an NVIDIA L40S by the hour in June 2026 is about $0.86/hr on RunPod. We track 8 providers publishing an on-demand price for it , and spot/community rates reach as low as $0.74/hr. Inference, fine-tuning, rendering and generative-image workloads. Strong FP8/FP16 throughput at a fraction of H100 cost; no NVLink.
Source: Provider pricing pages. Data as of June 2026.
NVIDIA L40S price by provider
| Provider | Type | On-demand /hr | Spot/community /hr | Notes |
|---|---|---|---|---|
| RunPod | specialist | $0.86/hr | — | Secure Cloud. |
| Verda (formerly DataCrunch) | specialist | $1.37/hr | — | On-demand. |
| Crusoe | specialist | $1.50/hr | — | On-demand. |
| Nebius | specialist | $1.55/hr | $0.74/hr | From $1.55 (AMD CPU host); reserved from $0.74. |
| Paperspace (DigitalOcean) | specialist | $1.57/hr | — | DigitalOcean GPU Droplet. |
| Vultr Cloud GPU | specialist | $1.67/hr | — | On-demand. |
| AWS EC2 | hyperscaler | $1.86/hr | $1.27/hr | g6e.xlarge (1 GPU) $1.861/hr. |
| CoreWeave | specialist | $2.25/hr | — | On-demand. |
| Google Cloud | hyperscaler | — | — | Not offered (G2 uses L4). |
| Hyperstack | specialist | — | — | Not offered (L40 non-S ~$1.00). |
| Lambda | specialist | — | — | Not offered. |
| Microsoft Azure | hyperscaler | — | — | Not offered (A10 NVads is the nearest). |
| Together AI | specialist | — | — | Not offered. |
| Vast.ai | marketplace | — | — | Listed but no reliable median; L40 (non-S) ~$0.47. |
Source: Provider pricing pages. Data as of June 2026.
Sorted cheapest on-demand first; "—" means the provider does not publish a price for this GPU (see the note). Snapshot June 2026 — cloud GPU prices change weekly; verify on the provider's pricing page before relying on a figure.
About the NVIDIA L40S
Inference, fine-tuning, rendering and generative-image workloads. Strong FP8/FP16 throughput at a fraction of H100 cost; no NVLink.
- VRAM: 48GB GDDR6
- Architecture: Ada Lovelace
- Released: 2023
- Cheapest on-demand: $0.86/hr on RunPod
- Cheapest spot/community: $0.74/hr on Nebius
Frequently asked questions
How much does it cost to rent an NVIDIA L40S per hour?
The cheapest on-demand NVIDIA L40S we track is $0.86/hr on RunPod, with 8 providers publishing a price. Spot/community rates go as low as $0.74/hr. Snapshot June 2026 — cloud GPU prices change weekly; verify on the provider's pricing page before relying on a figure.
Which provider is cheapest for the NVIDIA L40S?
RunPod at $0.86/hr is the cheapest on-demand option in our June 2026 snapshot. Specialist neoclouds and marketplaces usually beat the hyperscalers (AWS/GCP/Azure) by 2-4x on the same GPU.
What is the NVIDIA L40S good for?
Inference, fine-tuning, rendering and generative-image workloads. Strong FP8/FP16 throughput at a fraction of H100 cost; no NVLink. It has 48GB GDDR6 of VRAM and uses the Ada Lovelace architecture (released 2023).
Is spot or on-demand cheaper for the NVIDIA L40S?
Spot, community and interruptible tiers can cut the NVIDIA L40S price roughly in half — the cheapest discounted rate we track is $0.74/hr versus $0.86/hr on-demand. The trade-off is your instance can be reclaimed, so use spot for fault-tolerant or checkpointed jobs.
Related GPUs
- NVIDIA L4 (24GB GDDR6)
- NVIDIA RTX 6000 Ada (48GB GDDR6)
- NVIDIA RTX A6000 (48GB GDDR6)
Keep exploring
Source & accuracy
Figures are a dated snapshot from each provider's published pricing page (as of June 2026). Hyperscaler per-GPU prices are derived from 8-GPU instance prices (instance price / 8) and include bundled vCPU/RAM; marketplace (Vast.ai) figures are moving medians, not list prices. We do not invent numbers — where a value is unknown we show "—" with a reason. See our methodology. This is an informational comparison, not a quote.
Last updated: 2026-06-21