L40S pricing on gpu.ai
Every gpu.ai L40S offering we track, normalized to a per-GPU hourly rate and ranked against 41 providers with current L40S pricing.
The cheapest L40S we currently track is $0.380/hr on Lium.io.
gpu.ai L40S rates
Hourly prices per GPU. Multi-GPU rows show the per-GPU rate, not the instance total.
| Price / GPU-hr | Type | GPUs |
|---|---|---|
| $0.800 | On-demand | 1 |
L40S specifications
- VRAM
- 48GB
- Architecture
- Ada Lovelace
- CUDA Cores
- 18,176
- Tensor Cores
- 568
- Memory Bandwidth
- 864 GB/s
- TDP
- 350W
L40S on other providers
- Lium.io$0.380/hr
- Vultr$0.418/hr
- Novita AI$0.550/hr
- 1Legion$0.620/hr
- Vast.ai$0.734/hr
- Nebius$0.740/hr
- VoltageGPU$0.760/hr
- Runpod$0.790/hr
- IO.NET$0.840/hr
- Cudo Compute$0.870/hr
- Shadeform$0.880/hr
- Civo$0.890/hr
- Runcrate$0.890/hr
- Packet AI$0.920/hr
- Verda$0.922/hr
- Omega Gradient$0.930/hr
- Seeweb$0.948/hr
- UpCloud$0.950/hr
- Spheron$0.960/hr
- Sesterce$0.968/hr
- Massed Compute$0.970/hr
- CoreWeave$0.985/hr
- Hexgrid Cloud$1.00/hr
- FireVM$1.03/hr
- Fluence Cloud$1.05/hr
- QuantaCloud$1.09/hr
- Koyeb$1.20/hr
- Gcore$1.20/hr
- AceCloud$1.50/hr
- Crusoe$1.50/hr
- DigitalOcean$1.57/hr
- Scaleway$1.64/hr
- Hugging Face$1.80/hr
- OVHcloud$1.80/hr
- Amazon AWS$1.86/hr
- Modal$1.95/hr
- Lyceum$1.99/hr
- Together AI$2.10/hr
- Hinode$3.30/hr
- Oracle Cloud$3.50/hr
About gpu.ai
gpu.ai is a provider-agnostic cloud compute platform that aggregates capacity from multiple providers.
Frequently Asked Questions
How much does the L40S cost on gpu.ai?
gpu.ai lists the L40S from $0.800 per GPU-hour across 1 tracked configuration. Prices are collected daily — see the table above for the current rates.
Is gpu.ai the cheapest place to rent the L40S?
gpu.ai ranks #9 of 41 providers we track with current L40S pricing. The cheapest right now is Lium.io. Price is only one factor — region availability, network, and storage costs differ between providers.
Which L40S configurations does gpu.ai offer?
We currently track 1 L40S offering from gpu.ai, covering the GPU counts, pricing types, and regions listed in the table above. Every price is normalized to a per-GPU hourly rate so it can be compared across providers.