Prices updated 2026-05-09 — sourced from live provider data

CoreWeave L40S Pricing Options

All available L40S instance types on CoreWeave, sorted by lowest hourly rate.

Region Instance Type Price/hr GPUs vCPUs RAM
US L40S On-Demand $0.9900/hr Cheapest 1× GPU 8 vCPU 96 GB
US L40S_x1 On-Demand $3.2000/hr 1× GPU 16 vCPU 128 GB
US L40S_x8 On-Demand $25.6000/hr 8× GPU 128 vCPU 1024 GB

Top Alternatives to CoreWeave for L40S

Side-by-side cost comparison vs the cheapest CoreWeave L40S rate ($0.9900/hr).

Provider Price/hr Type vs CoreWeave
Vast.ai $0.4200/hr Spot -58% cheaper vs CoreWeave
FluidStack $0.5900/hr On-Demand -40% cheaper vs CoreWeave
TensorDock $0.5900/hr On-Demand -40% cheaper vs CoreWeave
RunPod $0.6700/hr Spot -32% cheaper vs CoreWeave
Jarvis Labs $0.6900/hr On-Demand -30% cheaper vs CoreWeave

View all L40S providers →

Calculate Your Exact L40S Cost

Enter your hours, job type, and GPU count. Our calculator compares CoreWeave against all 32 providers instantly.

Frequently Asked Questions

How much does CoreWeave L40S cost per hour in 2026? +
CoreWeave L40S starts at $0.9900/hr (on-demand). The average L40S on-demand price across all providers is $3.9322/hr. Use the GPU cost calculator to estimate total cost for your workload.
What are the cheapest alternatives to CoreWeave for L40S? +
Vast.ai at $0.4200/hr, FluidStack at $0.5900/hr, TensorDock at $0.5900/hr, and more. See the full comparison table for all providers and pricing types.
Does CoreWeave offer reserved or spot L40S pricing? +
CoreWeave offers the following L40S pricing types: on-demand. Reserved pricing typically saves 20–35% vs on-demand for long-running workloads.