GPU / INDEX / A100
The same A100 costs 3.8× more on some clouds than others.
The workhorse · 40/80GB HBM2e. Every provider's cheapest A100 in the sample, ranked cheapest-first. On-demand $/GPU/hr. Color is temperature: cool is cheap, white-hot is expensive.
A100 · Median $/GPU/hr
$1.69/HR
Range across providers
$0.67 — $3.40/HR
§ 01 / THE TABLE
Cheapest A100 per provider.
| # | Provider | Variant | Region | $/GPU/hr | Spot |
|---|---|---|---|---|---|
| 01 | Vast.ai | A100 | South Korea, KR, AS | $0.67 2× | — |
| 02 | PrimeIntellect | A100 | FI | $0.72 8× | — |
| 03 | Jarvis Labs | a100-sxm-40 | — | $0.89 | — |
| 04 | Seeweb | A100 | — | $0.99 | — |
| 05 | Thunder Compute | a100 | — | $1.09 | — |
| 06 | Hyperstack | A100-80G-PCIe | CANADA 1 | $1.35 | — |
| 07 | Massed Compute | a100 | — | $1.35 | — |
| 08 | Shadeform VIA SHADEFORM | A100-80GB | Omaha USA 2 | $1.35 8× | — |
| 09 | Spheron | a100 | — | $1.48 | — |
| 10 | TensorDock | a100-pcie-80 | — | $1.50 | — |
| 11 | Cudo | A100 | NO Kristiansand 1 | $1.53 | — |
| 12 | DataCrunch | a100-sxm-40 | — | $1.69 | — |
| 13 | Verda | A100 | FIN 02 | $1.69 | $0.85 |
| 14 | RunPod | A100-80GB | IN | $1.79 | — |
| 15 | Lambda | A100 | Europe Central 1 | $1.99 8× | — |
| 16 | Crusoe | a100-pcie-80 | — | $2.00 | — |
| 17 | Vultr | a100 | — | $2.40 | — |
| 18 | CoreWeave | a100 | — | $2.70 | — |
| 19 | AWS | a100-sxm-40 | — | $2.74 | — |
| 20 | Google Cloud | a100-sxm-40 | — | $2.93 | — |
| 21 | OVHcloud | A100-80GB | GRA | $3.07 | — |
| 22 | Paperspace | A100-80GB | East Coast (NY2) | $3.18 8× | — |
| 23 | Azure | a100-sxm-40 | — | $3.40 | — |
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 A100 market sits.
Vast.ai$0.67 2×
PrimeIntellect$0.72 8×
Jarvis Labs$0.89
Seeweb$0.99
Thunder Compute$1.09
Hyperstack$1.35
Massed Compute$1.35
Shadeform VIA SHADEFORM$1.35 8×
Spheron$1.48
TensorDock$1.50
Cudo$1.53
DataCrunch$1.69
Verda$1.69
RunPod$1.79
Lambda$1.99 8×
Crusoe$2.00
Vultr$2.40
CoreWeave$2.70
AWS$2.74
Google Cloud$2.93
OVHcloud$3.07
Paperspace$3.18 8×
Azure$3.40
$0.67 · COOL$3.40 · WHITE-HOT
| Vast.ai | $0.67 per GPU/hour (from a 2-GPU node price) |
| PrimeIntellect | $0.72 per GPU/hour (from a 8-GPU node price) |
| Jarvis Labs | $0.89 per GPU/hour |
| Seeweb | $0.99 per GPU/hour |
| Thunder Compute | $1.09 per GPU/hour |
| Hyperstack | $1.35 per GPU/hour |
| Massed Compute | $1.35 per GPU/hour |
| Shadeform (via Shadeform) | $1.35 per GPU/hour (from a 8-GPU node price) |
| Spheron | $1.48 per GPU/hour |
| TensorDock | $1.50 per GPU/hour |
| Cudo | $1.53 per GPU/hour |
| DataCrunch | $1.69 per GPU/hour |
| Verda | $1.69 per GPU/hour |
| RunPod | $1.79 per GPU/hour |
| Lambda | $1.99 per GPU/hour (from a 8-GPU node price) |
| Crusoe | $2.00 per GPU/hour |
| Vultr | $2.40 per GPU/hour |
| CoreWeave | $2.70 per GPU/hour |
| AWS | $2.74 per GPU/hour |
| Google Cloud | $2.93 per GPU/hour |
| OVHcloud | $3.07 per GPU/hour |
| Paperspace | $3.18 per GPU/hour (from a 8-GPU node price) |
| Azure | $3.40 per GPU/hour |
§ / QUOTES
Running A100? 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.
22 DIRECT PROVIDERS · MEDIAN $1.69/GPU/HR · SAMPLE 2026-11-13