Loading prices…
L40S rental prices
NVIDIA · Ada Lovelace · 48 GB GDDR6 · PCIe · 2023
26 providers · 88 configurations · prices as of Nov 11, 2026, 05:01 UTC
Lowest on-demand
$0.38per GPU·h · ≈ $277/month
at LiumMedian on-demand
$1.40/GPU·h
across 25 providers
Lowest spot
$0.56/GPU·h
Reserved from $0.73
FP16/BF16 cost
$1.05/PFLOP·h
#1 of 72 GPUs for value
Providers
26
88 configurations
Price history
Daily L40S prices per GPU-hour. The shaded period is reconstructed from public archives: the version history of SkyPilot's price catalogs, Internet Archive copies of providers' pricing pages and the open gpu-price-feed project (fewer providers than live data). After that, PetaPrice's own tracking.
Lowest on-demandMedian on-demandLowest spot
Price by provider
Cheapest On-demand L40S offer at each provider, per GPU-hour. Shorter and greener is cheaper.
Lium$0.38best
Infomaniak🇨🇭$0.71+87%
Latitude.sh🇺🇸$0.74+95%
RunPod🇺🇸$0.792.1×
EmpirioLabs AI🇺🇸$0.802.1×
GPU.ai🇺🇸$0.872.3×
Cudo Compute🇬🇧$0.882.3×
Seeweb🇮🇹$0.952.5×
Massed Compute🇺🇸$0.972.6×
UpCloud🇫🇮$1.193.1×
Koyeb🇫🇷$1.203.2×
Civo🇬🇧$1.293.4×
Gcore🇱🇺$1.403.7×
Crusoe🇺🇸$1.503.9×
Nebius🇳🇱$1.554.1×
DigitalOcean🇺🇸$1.574.1×
OVHcloud🇫🇷$1.574.1×
Sesterce🇫🇷$1.654.3×
Vultr🇺🇸$1.674.4×
Scaleway🇫🇷$1.704.5×
AWS🇺🇸$1.864.9×
STACKIT🇩🇪$1.955.1×
Verda (DataCrunch)🇫🇮$1.985.2×
CoreWeave🇺🇸$2.255.9×
Oracle Cloud🇺🇸$3.509.2×
Thin vertical line = median across providers ($1.40).
Every provider we track
Pricing
Size
Host
Provider HQ
26 providers · 88 configurations
Sort
- LiumcommunityOn-demandHQ unknownvia Liumfrom · On-demandCheapest$0.38/GPU·hCheapest
- 2× GPU23 vCPU · 141 GB RAM · 1,229 GB disk$0.38 /GPU·h$0.76 /hOn-demand🇺🇸 Las Vegas, US● availableRent ↗
- InfomaniakOn-demand🇨🇭 HQ in Switzerlandvia Infomaniak pricing pagefrom · On-demand+87%$0.71/GPU·h+87%
- 1× GPU16 vCPU · 32 GB RAM$0.71 /GPU·h$0.71 /hOn-demandSwitzerlandRent ↗
- Latitude.shOn-demand🇺🇸 HQ in United Statesvia Shadeformfrom · On-demand+95%$0.74/GPU·h+95%
- 1× GPU · PCIe16 vCPU · 128 GB RAM · 500 GB disk$0.74 /GPU·h$0.74 /hOn-demand🇩🇪 DE, Frankfurt○ sold outRent ↗
- RunPodcommunityOn-demand🇺🇸 HQ in United Statesvia RunPodfrom · On-demand2.1×$0.79/GPU·h2.1×
- EmpirioLabs AIOn-demand🇺🇸 HQ in United Statesvia EmpirioLabs APIfrom · On-demand2.1×$0.80/GPU·h2.1×
- 2× GPU23 vCPU · 137 GB RAM · 1,069 GB disk$0.80 /GPU·h$1.60 /hOn-demandUS● availableRent ↗
- GPU.aicommunityOn-demand🇺🇸 HQ in United Statesvia GPU.aifrom · On-demand2.3×$0.87/GPU·h2.3×
- Cudo ComputeOn-demand🇬🇧 HQ in United Kingdomvia Cudo APIfrom · On-demand2.3×$0.88/GPU·h2.3×
- 1× GPU1 vCPU · 1 GB RAM$0.88 /GPU·h$0.88 /hOn-demandno-kristiansand-1● availableRent ↗
- SeewebOn-demandReserved🇮🇹 HQ in Italyvia Seeweb pricing pagefrom · On-demand2.5×$0.95/GPU·h2.5×
What can the L40S run?
Memory for the weights plus one 32K-token sequence of context, how many GPUs hold it, and what that replica costs at the cheapest in-stock on-demand rate.
| Model | Params | Memory | GPUs | $/hr | $/mo |
|---|---|---|---|---|---|
gpt-oss-20b OpenAI · 20.9B | 20.9B · 3.6B active | 20 GiB | 1 | $0.38 at Lium | $277 |
Gemma 3 27B Google · 27.4B | 27.4B | 28 GiB | 1 | $0.38 at Lium | $277 |
Qwen3-32B Alibaba · 32.8B | 32.8B | 39 GiB | 1 | $0.38 at Lium | $277 |
Llama 3.3 70B Instruct Meta · 70.6B | 70.6B | 76 GiB | 2 | $0.76 at Lium | $555 |
gpt-oss-120b OpenAI · 117B | 117B · 5.1B active | 110 GiB | 4 | $1.52 at Lium | $1,110 |
Qwen3-235B-A22B Alibaba · 235B | 235B · 22B active | 225 GiB | 8 | $3.04 at Lium | $2,219 |
DeepSeek-R1 DeepSeek · 684B | 684B · 37B active | 640 GiB | 16 | $6.08 at Lium | $4,438 |
Kimi K3 Moonshot AI · 2.78T | 2.78T · 104B active | 2.53 TiB | 64 | $24.32 at Lium | $17,754 |
Assumes 90% of each GPU's memory is usable and rounds up to 1, 2, 4, 8… GPUs for tensor parallelism (up to 64). $/hr is the GPU count times the cheapest in-stock on-demand rate per GPU; check that the provider rents instances that large. Serving many requests at once needs more memory for context.
Alternatives to the L40S
GPUs with similar compute and memory, and their lowest on-demand price right now.
- A100 SXM 40GB40 GB · 0.312 PF FP16$0.91/h$2.92/PFLOP·h
- A100 PCIe 40GB40 GB · 0.312 PF FP16$1.09/h$3.49/PFLOP·h
- A100 SXM 80GB80 GB · 0.312 PF FP16$1.07/h$3.43/PFLOP·h
- A100 PCIe 80GB80 GB · 0.312 PF FP16$1.11/h$3.56/PFLOP·h
- RTX 6000 Ada48 GB · 0.364 PF FP16$0.63/h$1.73/PFLOP·h
Gaudi 296 GB · 0.432 PF FP16$1.10/h$2.55/PFLOP·h
L40S pricing questions
How much does it cost to rent an L40S?
On-demand L40S prices start at $0.38 per GPU-hour (Lium). The median across 25 providers is $1.40, and the most expensive is $3.50. Running one GPU around the clock costs about $277 per month at the lowest price.
Is there cheaper spot or interruptible L40S capacity?
Yes. Spot (interruptible) L40S capacity starts at $0.56 per GPU-hour, -47% 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 L40S?
At $0.38 per hour and 0.362 PF of dense FP16/BF16, the L40S costs $1.05 per PFLOP-hour. That ranks 1 of 72 GPUs with on-demand prices on PetaPrice. Figures are datasheet peaks; real utilization is typically 30–60%.
What are the L40S's key specs?
48 GB GDDR6 with 864 GB/s of memory bandwidth, Ada Lovelace architecture (2023), PCIe form factor, 350 W. Dense peak: 0.362 PF FP16 / BF16, 0.733 PF FP8, 0.733 PF INT8, 0.183 PF TF32, 0.0916 PF FP32, 0.0014 PF FP64.
How current are these prices?
Prices were collected on Nov 11, 2026, 05:01 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.