Live GPU pricing from 20+ providers  ·  Free to use
GPUHunt/Use Cases/Best GPU Servers for AI Inference

Best GPU Servers for AI Inference

Inference workloads need high throughput per dollar, not raw training speed. L40S, L4, and A10 hit the sweet spot between VRAM, compute, and cost. Compare live pricing across all providers.

87
Options
$0.34
From /hr
16
Providers
6
GPU types

What to look for

  • L40S offers 48 GB GDDR6 with Ada Lovelace architecture — 2× the inference throughput of A100 per dollar
  • L4 is purpose-built for inference: low power, 24 GB VRAM, runs 7B–13B models efficiently
  • A10 / A10G are workhorses for high-concurrency serving at a fraction of H100 cost
  • RTX 4090 is viable for small models and prototyping at consumer pricing
Recommended GPU families
87 results
Density:
RunPodRunPod Global
24 GB VRAM total
$0.34/hr
$0.014/GB·hr
View deal
RunPodRunPod Global
24 GB VRAM total
$0.44/hr
$0.018/GB·hr
View deal
Jarvis LabsJarvis Labs🇺🇸 US
8 cores60 GB RAM24 GB VRAM total
$0.49/hr
$0.020/GB·hr
View deal
TensorDockTensorDock🇺🇸 US
8 cores48 GB RAM24 GB VRAM total
$0.50/hr
$0.021/GB·hr
View deal
Lambda LabsLambda Labs🇺🇸 US
30 cores200 GB RAM24 GB VRAM total
$0.60/hr
$0.025/GB·hr
View deal
OblivusOblivus🇪🇺 EU
12 cores70 GB RAM24 GB VRAM total
$0.64/hr
$0.027/GB·hr
View deal
RunPodRunPod Global
48 GB VRAM total
$0.69/hr
$0.014/GB·hr
View deal
ScalewayScaleway🇫🇷 EU-FR
8 cores48 GB RAM24 GB VRAM total
$0.75/hr
$0.031/GB·hr
View deal
Thunder ComputeThunder Compute🇺🇸 US
4 cores24 GB RAM80 GB VRAM total
$0.78/hr
$0.010/GB·hr
View deal
RunPodRunPod Global
48 GB VRAM total
$0.79/hr
$0.016/GB·hr
View deal
FluidStackFluidStack🇪🇺 EU
8 cores48 GB RAM24 GB VRAM total
$0.80/hr
$0.033/GB·hr
View deal
OblivusOblivus🇪🇺 EU
28 cores58 GB RAM48 GB VRAM total
$1.05/hr
$0.022/GB·hr
View deal
Lambda LabsLambda Labs🇺🇸 US
30 cores200 GB RAM40 GB VRAM total
$1.10/hr
$0.028/GB·hr
View deal
RunPodRunPod Global
80 GB VRAM total
$1.19/hr
$0.015/GB·hr
View deal
FluidStackFluidStack🇺🇸 US
16 cores100 GB RAM48 GB VRAM total
$1.25/hr
$0.026/GB·hr
View deal
TensorDockTensorDock🇺🇸 US
16 cores120 GB RAM48 GB VRAM total
$1.29/hr
$0.027/GB·hr
View deal
HyperstackHyperstack🇺🇸 US
12 cores96 GB RAM48 GB VRAM total
$1.29/hr
$0.027/GB·hr
View deal
Lambda LabsLambda Labs🇺🇸 US
30 cores200 GB RAM80 GB VRAM total
$1.29/hr
$0.016/GB·hr
View deal
FluidStackFluidStack🇺🇸 US
16 cores117 GB RAM40 GB VRAM total
$1.35/hr
$0.034/GB·hr
View deal
TensorDockTensorDock🇺🇸 US
16 cores120 GB RAM40 GB VRAM total
$1.35/hr
$0.034/GB·hr
View deal
RunPodRunPod Global
80 GB VRAM total
$1.39/hr
$0.017/GB·hr
View deal
OblivusOblivus🇪🇺 EU
28 cores120 GB RAM80 GB VRAM total
$1.47/hr
$0.018/GB·hr
View deal
HyperstackHyperstack🇺🇸 US
16 cores120 GB RAM48 GB VRAM total
$1.49/hr
$0.031/GB·hr
View deal
ScalewayScaleway🇫🇷 EU-FR
12 cores96 GB RAM48 GB VRAM total
$1.49/hr
$0.031/GB·hr
View deal
ScalewayScaleway🇫🇷 EU-FR
16 cores96 GB RAM48 GB VRAM total
$1.50/hr
$0.031/GB·hr
View deal
DigitalOceanDigitalOcean🇺🇸 US
8 cores64 GB RAM48 GB VRAM total
$1.57/hr
$0.033/GB·hr
View deal
OblivusOblivus🇪🇺 EU
31 cores240 GB RAM80 GB VRAM total
$1.57/hr
$0.020/GB·hr
View deal
OVHcloudOVHcloud🇺🇸 US
192 GB RAM48 GB VRAM total
$1204/mo
View deal
CoreWeaveCoreWeave🇺🇸 US
16 cores120 GB RAM48 GB VRAM total
$1.69/hr
$0.035/GB·hr
View deal
OVHcloudOVHcloud🇺🇸 US
192 GB RAM48 GB VRAM total
$1241/mo
View deal
PaperspacePaperspace🇺🇸 US
8 cores45 GB RAM40 GB VRAM total
$1.71/hr
$0.043/GB·hr
View deal
OVHcloudOVHcloud🇺🇸 US
192 GB RAM48 GB VRAM total
$1279/mo
View deal
FluidStackFluidStack🇺🇸 US
16 cores117 GB RAM80 GB VRAM total
$1.79/hr
$0.022/GB·hr
View deal
CoreWeaveCoreWeave🇺🇸 US
24 cores192 GB RAM40 GB VRAM total
$1.79/hr
$0.045/GB·hr
View deal
HyperstackHyperstack🇪🇺 EU
16 cores120 GB RAM80 GB VRAM total
$1.79/hr
$0.022/GB·hr
View deal
Jarvis LabsJarvis Labs🇺🇸 US
12 cores120 GB RAM80 GB VRAM total
$1.79/hr
$0.022/GB·hr
View deal
TensorDockTensorDock🇺🇸 US
20 cores160 GB RAM80 GB VRAM total
$1.89/hr
$0.024/GB·hr
View deal
Jarvis LabsJarvis Labs🇺🇸 US
32 cores240 GB RAM96 GB VRAM total
$1.96/hr
$0.020/GB·hr
View deal
CoreWeaveCoreWeave🇺🇸 US
24 cores192 GB RAM80 GB VRAM total
$1.99/hr
$0.025/GB·hr
View deal
HyperstackHyperstack🇺🇸 US
20 cores180 GB RAM80 GB VRAM total
$1.99/hr
$0.025/GB·hr
View deal
TensorDockTensorDock🇺🇸 US
32 cores192 GB RAM96 GB VRAM total
$2.00/hr
$0.021/GB·hr
View deal
CoreWeaveCoreWeave🇺🇸 US
30 cores200 GB RAM80 GB VRAM total
$2.06/hr
$0.026/GB·hr
View deal
PaperspacePaperspace🇺🇸 US
12 cores90 GB RAM80 GB VRAM total
$2.30/hr
$0.029/GB·hr
View deal
DataCrunchDataCrunch🇫🇮 FI-01
22 cores184 GB RAM80 GB VRAM total
$2.56/hr
$0.032/GB·hr
View deal
Lambda LabsLambda Labs🇺🇸 US
60 cores400 GB RAM160 GB VRAM total
$2.58/hr
$0.016/GB·hr
View deal
ScalewayScaleway🇫🇷 EU-FR
24 cores192 GB RAM96 GB VRAM total
$2.98/hr
$0.031/GB·hr
View deal
Genesis CloudGenesis Cloud🇪🇺 EU
32 cores192 GB RAM80 GB VRAM total
$2.99/hr
$0.037/GB·hr
View deal
ScalewayScaleway🇫🇷 EU-FR
32 cores192 GB RAM96 GB VRAM total
$3.00/hr
$0.031/GB·hr
View deal
PaperspacePaperspace🇺🇸 US
12 cores90 GB RAM80 GB VRAM total
$3.18/hr
$0.040/GB·hr
View deal
Jarvis LabsJarvis Labs🇺🇸 US
24 cores240 GB RAM160 GB VRAM total
$3.58/hr
$0.022/GB·hr
View deal
Latitude.shLatitude.sh🇺🇸 US
32 cores256 GB RAM96 GB VRAM total
$4.00/hr
$0.042/GB·hr
View deal
OblivusOblivus🇪🇺 EU
112 cores232 GB RAM192 GB VRAM total
$4.20/hr
$0.022/GB·hr
View deal
Latitude.shLatitude.sh🇺🇸 US
64 cores512 GB RAM160 GB VRAM total
$4.48/hr
$0.028/GB·hr
View deal
OVHcloudOVHcloud🇺🇸 US
384 GB RAM96 GB VRAM total
$3505/mo
View deal
OVHcloudOVHcloud🇺🇸 US
384 GB RAM96 GB VRAM total
$3505/mo
View deal
OVHcloudOVHcloud🇺🇸 US
384 GB RAM96 GB VRAM total
$3505/mo
View deal
DataCrunchDataCrunch🇫🇮 FI-01
44 cores368 GB RAM160 GB VRAM total
$5.12/hr
$0.032/GB·hr
View deal
OblivusOblivus🇪🇺 EU
120 cores706 GB RAM192 GB VRAM total
$5.12/hr
$0.027/GB·hr
View deal
HyperstackHyperstack🇺🇸 US
48 cores384 GB RAM192 GB VRAM total
$5.16/hr
$0.027/GB·hr
View deal
Lambda LabsLambda Labs🇺🇸 US
120 cores800 GB RAM320 GB VRAM total
$5.16/hr
$0.016/GB·hr
View deal
Latitude.shLatitude.sh🇺🇸 US
64 cores512 GB RAM192 GB VRAM total
$5.60/hr
$0.029/GB·hr
View deal
ScalewayScaleway🇫🇷 EU-FR
48 cores384 GB RAM192 GB VRAM total
$5.96/hr
$0.031/GB·hr
View deal
ScalewayScaleway🇫🇷 EU-FR
64 cores384 GB RAM192 GB VRAM total
$6.00/hr
$0.031/GB·hr
View deal
FluidStackFluidStack🇪🇺 EU
64 cores384 GB RAM192 GB VRAM total
$6.40/hr
$0.033/GB·hr
View deal
HyperstackHyperstack🇪🇺 EU
64 cores480 GB RAM320 GB VRAM total
$7.16/hr
$0.022/GB·hr
View deal
Jarvis LabsJarvis Labs🇺🇸 US
48 cores480 GB RAM320 GB VRAM total
$7.16/hr
$0.022/GB·hr
View deal
Latitude.shLatitude.sh🇺🇸 US
64 cores512 GB RAM192 GB VRAM total
$8.00/hr
$0.042/GB·hr
View deal
OblivusOblivus🇪🇺 EU
252 cores464 GB RAM384 GB VRAM total
$8.40/hr
$0.022/GB·hr
View deal
FluidStackFluidStack🇺🇸 US
128 cores800 GB RAM384 GB VRAM total
$10.00/hr
$0.026/GB·hr
View deal
DataCrunchDataCrunch🇫🇮 FI-01
88 cores736 GB RAM320 GB VRAM total
$10.24/hr
$0.032/GB·hr
View deal
Lambda LabsLambda Labs🇺🇸 US
240 cores1800 GB RAM640 GB VRAM total
$10.32/hr
$0.016/GB·hr
View deal
FluidStackFluidStack🇪🇺 EU
128 cores936 GB RAM320 GB VRAM total
$10.80/hr
$0.034/GB·hr
View deal
Latitude.shLatitude.sh🇺🇸 US
128 cores1024 GB RAM384 GB VRAM total
$11.20/hr
$0.029/GB·hr
View deal
Latitude.shLatitude.sh🇺🇸 US
128 cores1024 GB RAM640 GB VRAM total
$11.20/hr
$0.017/GB·hr
View deal
OblivusOblivus🇪🇺 EU
252 cores1440 GB RAM640 GB VRAM total
$11.76/hr
$0.018/GB·hr
View deal
HyperstackHyperstack🇺🇸 US
128 cores960 GB RAM384 GB VRAM total
$11.92/hr
$0.031/GB·hr
View deal
ScalewayScaleway🇫🇷 EU-FR
96 cores768 GB RAM384 GB VRAM total
$11.92/hr
$0.031/GB·hr
View deal
OblivusOblivus🇪🇺 EU
252 cores1920 GB RAM640 GB VRAM total
$12.56/hr
$0.020/GB·hr
View deal
CoreWeaveCoreWeave🇺🇸 US
128 cores960 GB RAM384 GB VRAM total
$13.52/hr
$0.035/GB·hr
View deal
FluidStackFluidStack🇺🇸 US
128 cores936 GB RAM640 GB VRAM total
$14.32/hr
$0.022/GB·hr
View deal
Jarvis LabsJarvis Labs🇺🇸 US
96 cores960 GB RAM640 GB VRAM total
$14.32/hr
$0.022/GB·hr
View deal
Thunder ComputeThunder Compute🇺🇸 US
32 cores192 GB RAM640 GB VRAM total
$14.32/hr
$0.022/GB·hr
View deal
TensorDockTensorDock🇺🇸 US
160 cores1280 GB RAM640 GB VRAM total
$15.12/hr
$0.024/GB·hr
View deal
HyperstackHyperstack🇺🇸 US
160 cores1440 GB RAM640 GB VRAM total
$15.92/hr
$0.025/GB·hr
View deal
CoreWeaveCoreWeave🇺🇸 US
240 cores1600 GB RAM640 GB VRAM total
$16.48/hr
$0.026/GB·hr
View deal
PaperspacePaperspace🇺🇸 US
96 cores720 GB RAM640 GB VRAM total
$18.40/hr
$0.029/GB·hr
View deal
DataCrunchDataCrunch🇫🇮 FI-01
176 cores1472 GB RAM640 GB VRAM total
$20.48/hr
$0.032/GB·hr
View deal