GPU MODELS / L40S

L40S cloud GPU pricing

74 listings across 48 providers. Cheapest published rate: $0.250/GPU/hr at SpotGPUs. Sorted by per-GPU hourly price. Every row links to the provider's own published price.

WHAT IS THE L40S

The L40S is the workhorse of the inference-and-fine-tuning middle: Ada tensor cores with FP8, 48 GB of memory, and a price well below the H100 line. It has no NVLink, so multi-GPU scaling is PCIe-limited - but for single-card and small-fleet serving, it is the most common answer on the panel, and often the best $/token in the catalog.

MAKER
NVIDIA
ARCHITECTURE
Ada Lovelace
LAUNCHED
2023
MEMORY
48 GB GDDR6 ECC
MEMORY BANDWIDTH
864 GB/s
TENSOR COMPUTE
362 TFLOPS FP16/BF16 tensor dense (733 sparse); 733 TFLOPS FP8 dense
POWER DRAW
350 W
INTERCONNECT
PCIe 4.0 x16 (no NVLink)

WHAT IT IS GOOD AT

LLM inference up to ~34B
48 GB serves quantized 34B or BF16 13B per card with FP8 throughput to spare.
Fine-tuning 7B-13B
LoRA and full fine-tunes of small models run economically; 864 GB/s GDDR6 is the limit for memory-hungry jobs.
Image and video generation
Ada RT and tensor cores suit diffusion workloads; 48 GB handles large batch image pipelines.

WHO SHOULD NOT RENT THE L40S

70B-class training or serving - 48 GB GDDR6 and no NVLink are hard limits.
Pure raster/RT graphics pipelines - the L40 is tuned for that and usually costs less.
ProviderConfigVRAMLocation$/hr$/GPU/hrSource
SpotGPUsSpotGPUsL40S × 148 GBNot published$0.250$0.250receipt
Novita AINovita AIL40S × 148 GBNot published$0.550$0.550receipt
AIC CloudAIC CloudL40S × 148 GBMumbai$0.563$0.563receipt
CloudRiftCloudRiftL40S × 148 GBNot published$0.570$0.570receipt
Baku ComputeBaku ComputeL40S (from) × 848 GBAzerbaijan (Baku)$5.200$0.650receipt
HydraHostHydraHostL40S × 148 GBNot published$0.700$0.700receipt
BeamBeamL40S (from) × 148 GBNot published$0.760$0.760receipt
RunPodRunPodL40S × 148 GBGlobal$0.790$0.790receipt
IO.NETIO.NETL40S × 148 GBNot published$0.790$0.790receipt
Applications.kzApplications.kzL40S (from) × 148 GBKazakhstan$0.790$0.790receipt
gpu.aigpu.aiL40S × 148 GBNot published$0.800$0.800receipt
VultrVultrL40S (bare metal) × 848 GBNot published$6.784$0.848receipt
InHostedInHostedL40S × 148 GBIndia$0.866$0.866receipt
Hostnot GPUHostnot GPUL40S × 148 GBAsia-Pacific$0.870$0.870receipt
Hostnot GPUHostnot GPUL40S × 148 GBUS Central$0.880$0.880receipt
Hostnot GPUHostnot GPUL40S × 248 GBUS Central$1.760$0.880receipt
Hostnot GPUHostnot GPUL40S × 448 GBUS Central$3.520$0.880receipt
Hostnot GPUHostnot GPUL40S × 848 GBUS Central$7.040$0.880receipt
packet.aipacket.aiL40S × 148 GBUS East · Virginia$0.920$0.920receipt
SpheronSpheronL40S × 148 GBNot published$0.960$0.960receipt
SeewebSeewebL40S × 148 GBItaly$0.969$0.969receipt
Massed ComputeMassed ComputeL40S × 148 GBNot published$0.970$0.970receipt
Massed ComputeMassed ComputeL40S × 848 GBNot published$7.760$0.970receipt
HexgridHexgridL40S (from) × 148 GBNot published$1.000$1.000receipt
E2E NetworksE2E NetworksL40S × 148 GBIndia$1.064$1.064receipt
LyceumLyceumL40S × 148 GBEurope$1.190$1.190receipt
KoyebKoyebL40S × 148 GBNot published$1.200$1.200receipt
Hostnot GPUHostnot GPUL40S × 148 GBAsia-Pacific$1.200$1.200receipt
GcoreGcoreL40S (from) × 148 GBNot published$1.231$1.231receipt
UpCloudUpCloudL40S × 148 GBNot published$1.265$1.265receipt
CivoCivoL40S × 148 GBNot published$1.290$1.290receipt
CivoCivoL40S × 248 GBNot published$2.580$1.290receipt
CivoCivoL40S × 448 GBNot published$5.160$1.290receipt
CivoCivoL40S × 848 GBNot published$10.320$1.290receipt
CyfutureCyfutureL40S × 148 GBIndia$1.293$1.293receipt
GPUBrazilGPUBrazilL40S × 145 GBAsia-Pacific$1.300$1.300receipt
GPUBrazilGPUBrazilL40S × 145 GBEurope$1.300$1.300receipt
DCPDCPL40S (from) × 148 GBSaudi Arabia$1.387$1.387receipt
Shakti CloudShakti CloudL40S × 148 GBIndia (Navi Mumbai)$1.429$1.429receipt
Shakti CloudShakti CloudL40S × 296 GBIndia (Navi Mumbai)$2.868$1.434receipt
GPUBrazilGPUBrazilL40S × 148 GBGlobal$1.444$1.444receipt
TensorPoolTensorPoolL40S × 148 GBNot published$1.490$1.490receipt
CrusoeCrusoeL40S × 148 GBUS / Iceland$1.500$1.500receipt
Verda (formerly DataCrunch)Verda (formerly DataCrunch)L40S × 848 GBEurope (Finland / Iceland)$12.220$1.528receipt
Verda (formerly DataCrunch)Verda (formerly DataCrunch)L40S × 148 GBEurope (Finland / Iceland)$1.528$1.528receipt
Verda (formerly DataCrunch)Verda (formerly DataCrunch)L40S × 248 GBEurope (Finland / Iceland)$3.056$1.528receipt
Verda (formerly DataCrunch)Verda (formerly DataCrunch)L40S × 448 GBEurope (Finland / Iceland)$6.112$1.528receipt
NebiusNebiusL40S with Intel CPU × 148 GBEU / US$1.550$1.550receipt
DigitalOceanDigitalOceanL40S × 148 GBNot published$1.570$1.570receipt
VultrVultrL40S × 148 GBNot published$1.671$1.671receipt
M
Machine0
L40S × 148 GBNot published$1.727$1.727receipt
Hugging FaceHugging FaceL40S × 148 GBUS$1.800$1.800receipt
OVHcloudOVHcloudL40S (from) × 148 GBNot published$1.800$1.800receipt
NebiusNebiusL40S with AMD CPU × 148 GBEU / US$1.820$1.820receipt
AWSAWSL40S × 148 GBUS East (N. Virginia)$1.861$1.861receipt
NeysaNeysaL40S (from) × 148 GBIndia$1.950$1.950receipt
ModalModalL40S × 148 GBNot published$1.951$1.951receipt
CerebriumCerebriumL40S × 148 GBNot published$1.951$1.951receipt
Shakti CloudShakti CloudL40S (bare metal) × 4192 GBIndia (Navi Mumbai)$8.219$2.055receipt
Hugging FaceHugging FaceL40S × 448 GBUS$8.300$2.075receipt
Lightning AILightning AIL40S × 148 GBUS$2.140$2.140receipt
CoreWeaveCoreWeaveL40S × 848 GBNorth America / Europe$18.000$2.250receipt
Catalyst CloudCatalyst CloudL40S × 148 GBNew Zealand$2.339$2.339receipt
Catalyst CloudCatalyst CloudL40S × 248 GBNew Zealand$4.678$2.339receipt
Catalyst CloudCatalyst CloudL40S × 448 GBNew Zealand$9.357$2.339receipt
AWSAWSL40S × 448 GBUS East (N. Virginia)$10.493$2.623receipt
Hugging FaceHugging FaceL40S × 848 GBUS$23.500$2.938receipt
HinodeHinodeL40S (from) × 148 GBNot published$3.300$3.300receipt
Oracle OCIOracle OCIL40S × 148 GBNot published (global list)$3.500$3.500receipt
ReplicateReplicateL40S × 148 GBGlobal$3.510$3.510receipt
ReplicateReplicateL40S × 248 GBGlobal$7.020$3.510receipt
ReplicateReplicateL40S × 448 GBGlobal$14.040$3.510receipt
ReplicateReplicateL40S × 848 GBGlobal$28.080$3.510receipt
CTYunCTYunL40S × 148 GBChina$4.648$4.648receipt

THE L40S MARKET, BY THE NUMBERS

PROVIDERS
48
CONFIGURATIONS
74
MEDIAN $/GPU/HR
$1.39
SPREAD
18.6x

48 providers publish 74 L40S configurations on the panel today. Published per-GPU hourly rates run from $0.250 (SpotGPUs) to $4.65 (CTYun), a 19x spread between the cheapest and the most expensive published rate for the same chip. The median listing sits at $1.39/GPU/hr. 41 rows were reported in stock by the provider at last observation; the rest are published catalog rates. All rates are on-demand - committed-use and reserved pricing is excluded.

CHEAPEST L40S BY GEOGRAPHY
indiaAIC Cloud$0.563/hrunited statesHostnot GPU$0.880/hritalySeeweb$0.969/hricelandCrusoe$1.500/hrfinlandVerda (formerly DataCrunch)$1.528/hrnew zealandCatalyst Cloud$2.339/hrchinaCTYun$4.648/hr
L40S PRICE TREND, DAILY PANEL SNAPSHOTS

Across every L40S row on the panel, the daily floor moved from $0.790 to $0.250/GPU/hr over 11 days of tracking (-68.4%). The median continuously-tracked listing moved from $1.44 to $1.50 (+4.2%). Snapshots run daily; the full history is public in the repo.

L40S PROVIDER NOTES
SpotGPUs · 1 config · 1 in stock
Sets the panel floor for this model.
from $0.250/hr
Novita AI · 1 config
Published catalog rates.
from $0.550/hr
AIC Cloud · 1 config · 1 region · 1 in stock
1 of 1 configs reported in stock at last check.
from $0.563/hr
CloudRift · 1 config
Published catalog rates.
from $0.570/hr
Baku Compute · 1 config · 1 region
Published catalog rates.
from $0.650/hr
HydraHost · 1 config · 1 in stock
1 of 1 configs reported in stock at last check.
from $0.700/hr
Beam · 1 config
Published catalog rates.
from $0.760/hr
RunPod · 1 config · 1 region · 1 in stock
1 of 1 configs reported in stock at last check.
from $0.790/hr
BUYING L40S: WHAT THE PANEL SAYS

At the current floor, one L40S running around the clock costs about $183 a month (730 hours of on-demand arithmetic). Node shape matters: the single-GPU floor is $0.250/GPU/hr against $0.650/GPU/hr on an 8-GPU node (+160.0%). Per-GPU floors by config size: 1x $0.25 · 2x $0.88 · 4x $0.88 · 8x $0.65. Compare the alternatives below before committing - the same budget often buys more than one chip class.

L40S VS THE ALTERNATIVES

L40S vs A100 80GB
L40S has more FP16 compute on paper, but GDDR6 bandwidth is less than half the A100's HBM2e and 48 GB caps model size. For 70B-class work the A100 wins; for smaller models the L40S is usually cheaper per token.
L40S vs L40
Same card with graphics-first tuning: the L40S trades some RT/raster throughput for roughly double the tensor compute. For AI work, always the L40S; for rendering, the L40.
L40S vs H100 PCIe
H100 PCIe adds HBM bandwidth, FP8 maturity and 80 GB - at 2-4x the hourly rate. L40S owns the middle market.

Compare live rates: L40S vs A100 80GB SXM · L40S vs A100 80GB · L40S vs L40 · L40S vs H100 PCIe · L40S vs L4 · L40S vs A40 · L40S vs RTX A6000 · L40S vs RTX 6000 Ada · L40S vs RTX 4090 · L40S vs RTX 5000 Ada · L40S vs L20 · L40S vs H100 SXM

L40S PRICING FAQ

What is the cheapest L40S cloud GPU right now?
The cheapest published L40S rate on the panel is $0.250 per GPU-hour at SpotGPUs (L40S x 1, Not published). Every price links to the provider's own published rate.
How much does a L40S cost per hour?
Across 48 providers tracked today, published per-GPU hourly rates for L40S run from $0.25 to $4.65, with a median of $1.39. Committed-use and reserved rates are excluded; everything shown is on-demand.
How many providers rent L40S GPUs?
48 providers publish 74 distinct L40S configurations on Compute Cafe. 41 of those rows were reported in stock by the provider's live endpoint at last observation.
Where is L40S cheapest?
By published per-GPU hourly rate, the cheapest geography for L40S today is india at $0.56/hr (AIC Cloud). The spread across 7 tracked geographies runs to $4.65/hr in china.
Are L40S prices going up or down?
Over the 11 days Compute Cafe has tracked the panel, the L40S floor is falling: $0.790 to $0.250/GPU/hr (-68.4%). The median continuously-tracked listing moved from $1.44 to $1.50 (+4.2%). Daily snapshots are public in the repo, so the trend is auditable.
What does a L40S cost per month?
At the current floor of $0.250/GPU/hr, a L40S running 24/7 costs about $183 per month (730 hours of on-demand arithmetic on the cheapest published rate). Committed-use contracts can price lower; the panel tracks on-demand rates only.
Is a full 8-GPU L40S node cheaper than a single card?
On today's panel: the single-GPU floor is $0.250/GPU/hr and the 8-GPU node floor is $0.650/GPU/hr - more expensive per GPU on a full node (+160.0%). Check the table for the exact configs behind each floor.
Is the L40S good for LLM inference?
It is the panel's default mid-market serving chip: models up to ~34B quantized run well, FP8 support keeps throughput high, and hourly rates are far below Hopper. Above that size, memory bandwidth becomes the constraint.
L40S vs A100 for fine-tuning?
For 7B-13B LoRA work the L40S is typically cheaper per hour and fast enough. For 70B-class or memory-bandwidth-bound jobs, the A100 80GB's HBM2e wins regardless of list compute.

See also alternatives to the L40S · cheapest GPU-hour per model · inference chips · training chips · by location