L40 rental prices: cheapest $/hr today
The cheapest on-demand L40 rental today is $0.79/hr on Thunder Compute(on-demand): verified 2026-08-22. 48 GB VRAM, NVIDIA.
Pricing at a glance
All current L40 offers
| Provider | Tier | $/hr | Verified | Source | |
|---|---|---|---|---|---|
| RunPod | community cloud | $0.69 | 2026-08-22 | source ↗ | Rent → |
| Thunder Compute | on-demand | $0.79 | 2026-08-22 | source ↗ | Rent → |
| RunPod | secure cloud | $0.82 | 2026-08-22 | source ↗ | Rent → |
| Massed Compute | on-demand | $0.86 | 2026-08-22 | source ↗ | Rent → |
| TensorDock | on-demand | $0.95 | 2026-08-22 | source ↗ | Rent → |
| Hyperstack | on-demand | $1.00 | 2026-08-22 | source ↗ | Rent → |
| CoreWeave 8× min | on-demand | $1.25 | 2026-08-22 | source ↗ | Rent → |
Tables rank by price, lowest first. Commercial relationships never affect ordering.How we collect and rank prices.
Secure/on-demand = reserved datacenter capacity. Community/spot = cheaper but interruptible. Prices are per GPU per hour, before storage/egress. Every row shows when it was verified at its source.
L40 hardware specs
| Architecture | Ada Lovelace |
|---|---|
| VRAM | 48 GB |
| FP16 TFLOPS (dense) | 181 |
| Memory bandwidth | 864 GB/s |
| TDP | 300 W |
| Interconnect | PCIe Gen4 |
| Launched | 2022 |
Datasheet spec for the L40 variant. Source: www.nvidia.com.
Price history
on-demand rate per day · median on-demand rate per day (dashed), since 2026-07-05: $0.79 low · $0.82 high. Full ledger on the history hub.
FAQ
The cheapest L40 rental we have is $0.79 per GPU-hour on Thunder Compute (on-demand), verified 2026-08-22.
48 GB.
Thunder Compute has the cheapest verified on-demand rate today at $0.79 per GPU-hour; interruptible spot/community capacity starts at $0.69. Verified 2026-08-22.
Related
All 48 GB+ GPUs · GPUs under $1/hr · Price movers this week · Thunder Compute review
Should you rent or buy?
A rough rule: divide the hardware price by the rental rate to get break-even hours. Run the exact math (with electricity) in the rent-vs-buy calculator.