NVIDIA A100 GPU cloud prices
Previous-generation workhorse — more affordable, widely available.
Compare vs other GPUs →Not enough price history yet — the trend fills in as we log more.
Provider-balanced managed-cloud median per GPU-hour (USD) · live data
Key Specifications
ArchitectureAmpere
Memory40 / 80GB HBM2e
Memory Bandwidth1.6 – 2.0 TB/s
ReleaseQ2 2020
Compare Cloud Provider Prices
We’re tracking 42 providers for the NVIDIA A100 — on-demand runs $0.27 to $8.19/hr per GPU, with interruptible spot from $0.48/hr.
| Provider | Instance | Location | VRAM | Billing | $/GPU/hr | Availability | |
|---|---|---|---|---|---|---|---|
GOGPU OutletPublic pricing feed | 1× NVIDIA A100 PCIe | US | 40GB | STANDARD | $0.27 | Not reported | View ↗ |
LILium.ioPublic pricing feed | 1× NVIDIA A100 PCIe | Boydton, United States | 40GB | STANDARD | $0.3 | Not reported | View ↗ |
GOGPU OutletPublic pricing feed | 1× NVIDIA A100 SXM | US | 80GB | STANDARD | $0.34 | Not reported | View ↗ |
GOGPU OutletPublic pricing feed | 1× NVIDIA A100 SXM | CA | 80GB | STANDARD | $0.38 | Not reported | View ↗ |
 AWSPublic pricing feed | 8× NVIDIA A100 | eu-north-1 | 40GB | BURST · SPOT | $0.39Modeled | Not reported | View ↗ |
SISimplePodMarketplace | 1× NVIDIA A100 SXM | Location not disclosed | 40GB | STANDARD | $0.48 | Available | View ↗ |
SISimplePodMarketplace | 1× NVIDIA A100 SXM | Location not disclosed | 40GB | BURST · SPOT | $0.48 | Available | View ↗ |
 VerdaProvider API | 1× NVIDIA A100 SXM | Location not disclosed | 40 / 80GB HBM2e | BURST · SPOT | $0.645 | Not reported | View ↗ |
| VerdaPublic pricing feed | 1× NVIDIA A100 | FIN-01 | 40GB | BURST · SPOT | $0.645 | Not reported | View ↗ |
| VerdaPublic pricing feed | 8× NVIDIA A100 | FIN-02 | 40GB | BURST · SPOT | $0.645 | Not reported | View ↗ |
| VerdaPublic pricing feed | 1× NVIDIA A100 | FIN-03 | 40GB | BURST · SPOT | $0.645 | Not reported | View ↗ |
| VerdaPublic pricing feed | 8× NVIDIA A100 | FIN-01 | 40GB | BURST · SPOT | $0.645 | Not reported | View ↗ |
| VerdaPublic pricing feed | 8× NVIDIA A100 PCIe | Location not disclosed | 40GB | BURST · SPOT | $0.645 | Not reported | View ↗ |
| VerdaPublic pricing feed | 8× NVIDIA A100 | FIN-03 | 40GB | BURST · SPOT | $0.645 | Not reported | View ↗ |
| VerdaPublic pricing feed | 1× NVIDIA A100 PCIe | Location not disclosed | 40GB | BURST · SPOT | $0.645 | Not reported | View ↗ |
| VerdaPublic pricing feed | 1× NVIDIA A100 | FIN-02 | 40GB | BURST · SPOT | $0.645 | Not reported | View ↗ |
| VerdaProvider API | 8× NVIDIA A100 SXM | Location not disclosed | 40 / 80GB HBM2e | BURST · SPOT | $0.645 | Not reported | View ↗ |
GOGPU OutletPublic pricing feed | 1× NVIDIA A100 SXM | SI | 80GB | STANDARD | $0.67 | Not reported | View ↗ |
 AzurePublic pricing feed | 2× NVIDIA A100 | westus2 | 80GB | BURST · SPOT | $0.679Modeled | Not reported | View ↗ |
| AzurePublic pricing feed | 4× NVIDIA A100 | eastus | 80GB | BURST · SPOT | $0.679Modeled | Not reported | View ↗ |
| AzurePublic pricing feed | 4× NVIDIA A100 | westus3 | 80GB | BURST · SPOT | $0.679Modeled | Not reported | View ↗ |
| AzurePublic pricing feed | 1× NVIDIA A100 | eastus | 80GB | BURST · SPOT | $0.679Modeled | Not reported | View ↗ |
| AzurePublic pricing feed | 1× NVIDIA A100 | westus2 | 80GB | BURST · SPOT | $0.679Modeled | Not reported | View ↗ |
| AzurePublic pricing feed | 2× NVIDIA A100 | eastus2 | 80GB | BURST · SPOT | $0.679Modeled | Not reported | View ↗ |
| AzurePublic pricing feed | 1× NVIDIA A100 | westus3 | 80GB | BURST · SPOT | $0.679Modeled | Not reported | View ↗ |
| AzurePublic pricing feed | 4× NVIDIA A100 | westus2 | 80GB | BURST · SPOT | $0.679Modeled | Not reported | View ↗ |
| AzurePublic pricing feed | 4× NVIDIA A100 | eastus2 | 80GB | BURST · SPOT | $0.679Modeled | Not reported | View ↗ |
| AzurePublic pricing feed | 2× NVIDIA A100 | westus3 | 80GB | BURST · SPOT | $0.679Modeled | Not reported | View ↗ |
| AzurePublic pricing feed | 2× NVIDIA A100 | eastus | 80GB | BURST · SPOT | $0.679Modeled | Not reported | View ↗ |
| AzurePublic pricing feed | 1× NVIDIA A100 | eastus2 | 80GB | BURST · SPOT | $0.679Modeled | Not reported | View ↗ |
 CivoPublic pricing feed | 1× NVIDIA A100 PCIe | Location not disclosed | 40GB | COMMITTED | $0.69 | Not reported | View ↗ |
| CivoPublic pricing feed | 2× NVIDIA A100 PCIe | Location not disclosed | 40GB | COMMITTED | $0.69 | Not reported | View ↗ |
| CivoPublic pricing feed | 4× NVIDIA A100 PCIe | Location not disclosed | 40GB | COMMITTED | $0.69 | Not reported | View ↗ |
| CivoPublic pricing feed | 8× NVIDIA A100 PCIe | Location not disclosed | 40GB | COMMITTED | $0.69 | Not reported | View ↗ |
| AzurePublic pricing feed | 8× NVIDIA A100 | eastus | 40GB | BURST · SPOT | $0.738Modeled | Not reported | View ↗ |
 1LegionPublic pricing feed | 8× NVIDIA A100 SXM | Location not disclosed | 80GB | COMMITTED | $0.74 | Not reported | View ↗ |
GOGPU OutletPublic pricing feed | 1× NVIDIA A100 SXM | HR | 80GB | STANDARD | $0.74 | Not reported | View ↗ |
| AzurePublic pricing feed | 8× NVIDIA A100 | westus2 | 40GB | BURST · SPOT | $0.748Modeled | Not reported | View ↗ |
| AzurePublic pricing feed | 2× NVIDIA A100 | centralus | 80GB | BURST · SPOT | $0.767Modeled | Not reported | View ↗ |
| AzurePublic pricing feed | 1× NVIDIA A100 | centralus | 80GB | BURST · SPOT | $0.767Modeled | Not reported | View ↗ |
Frequently Asked Questions
Why choose the NVIDIA A100?
40/80GB HBM2e at a fraction of H100 pricing; plenty for fine-tuning and mid-size training/inference.
When is the NVIDIA A100 not a good fit?
No FP8 and lower throughput than Hopper — H100 wins for the largest training runs.
How much does the NVIDIA A100 cost per hour?
NVIDIA A100 on-demand pricing currently runs $0.27/hr to $8.19/hr per GPU, depending on the provider and instance type. Interruptible spot capacity starts around $0.48/hr.
How much does the NVIDIA A100 cost per month?
At 720 hours per month, one NVIDIA A100 runs about $194 to $5,899 per month on demand.
Which cloud providers offer the NVIDIA A100?
We currently track 1194 NVIDIA A100 listings across 42 providers. Pricing and availability vary by region and billing model.