GPU / INDEX / L40
The same L40 costs 1.6× more on some clouds than others.
48GB Ada · datacenter. Every provider's cheapest L40 in the sample, ranked cheapest-first. On-demand $/GPU/hr. Color is temperature: cool is cheap, white-hot is expensive.
L40 · Median $/GPU/hr
$0.87/HR
Range across providers
$0.79 — $1.25/HR
§ 01 / THE TABLE
Cheapest L40 per provider.
| # | Provider | Variant | Region | $/GPU/hr | Spot |
|---|---|---|---|---|---|
| 01 | Thunder Compute | l40 | — | $0.79 | — |
| 02 | RunPod | L40 | IN | $0.82 10× | — |
| 03 | Massed Compute | l40 | — | $0.86 | — |
| 04 | PrimeIntellect | L40 | DE | $0.87 8× | — |
| 05 | TensorDock | l40 | — | $0.95 | — |
| 06 | Hyperstack | L40 | CANADA 1 | $1.00 | — |
| 07 | Shadeform VIA SHADEFORM | L40 | Montreal Canada 2 | $1.00 | — |
| 08 | CoreWeave | l40 | — | $1.25 | — |
SOURCE: SKYPILOT CATALOG · CATALOGS/V8 · SAMPLED 2026-11-13 · PER-GPU-HR = LIST PRICE ÷ GPU COUNT
LIVE PIPELINE · 18 CLOUDS
§ 02 / MARKET SHAPE
Where the L40 market sits.
Thunder Compute$0.79
RunPod$0.82 10×
Massed Compute$0.86
PrimeIntellect$0.87 8×
TensorDock$0.95
Hyperstack$1.00
Shadeform VIA SHADEFORM$1.00
CoreWeave$1.25
$0.79 · COOL$1.25 · WHITE-HOT
| Thunder Compute | $0.79 per GPU/hour |
| RunPod | $0.82 per GPU/hour (from a 10-GPU node price) |
| Massed Compute | $0.86 per GPU/hour |
| PrimeIntellect | $0.87 per GPU/hour (from a 8-GPU node price) |
| TensorDock | $0.95 per GPU/hour |
| Hyperstack | $1.00 per GPU/hour |
| Shadeform (via Shadeform) | $1.00 per GPU/hour |
| CoreWeave | $1.25 per GPU/hour |
§ / QUOTES
Running L40? Get matched quotes.
Tell us the count, duration, and region. We review every submission by hand and connect you with providers that fit — expect a reply within one business day. No spam, no obligation.
7 DIRECT PROVIDERS · MEDIAN $0.87/GPU/HR · SAMPLE 2026-11-13