Best Cloud GPU for LLM inference 70B
Minimum 80GB VRAM · Recommended 160GB+ · Runtime: tokens-per-second
Cheapest for LLM inference 70B: NVIDIA A100 80GB SXM on Paperspace
$1.15/hr · verify on provider site
Cheapest GPU Options — 9 eligible GPUs
Some links are affiliate links — we may earn a commission at no cost to you.Details.
| Provider | Price/hr | Configuration | Region | Billing | Availability | Rent |
|---|---|---|---|---|---|---|
| Paperspacecheapest | $1.15/hr | 1x A100-80G | us-east-1 | per-hour | on-demand | Rent → |
| Paperspace | $1.15/hr | 1x A100-80G | us-east-1 | per-hour | on-demand | Rent → |
| Vast.ai | $1.59/hr | 1x A100 PCIe 80GB | us-east | per-second | on-demand | Rent → |
| RunPod | $1.79/hr | 1x A100 SXM | us-east | per-second | on-demand | Rent → |
| RunPod | $1.79/hr | 1x A100 SXM | us-east | per-second | on-demand | Rent → |
| FluidStack | $1.85/hr | 1x A100 SXM 80GB | us-east | per-minute | on-demand | Rent → |
| Vast.ai | $1.89/hr | 1x H100 PCIe | us-west | per-second | on-demand | Rent → |
| DataCrunch | $1.89/hr | 1x A100 SXM4 80GB | eu-north | per-minute | on-demand | Rent → |
| Genesis Cloud | $1.89/hr | 1x A100 PCIe 80GB | us-east | per-minute | on-demand | Rent → |
| TensorDock | $1.99/hr | 1x A100 PCIe 80GB | us-east | per-minute | on-demand | Rent → |
| Hyperstack | $2.06/hr | 1x A100 SXM4 80GB | uk-london | per-minute | on-demand | Rent → |
| FluidStack | $2.15/hr | 1x H100 PCIe | eu-west | per-minute | on-demand | Rent → |
| CoreWeave | $2.21/hr | 1x A100 SXM4 80GB | us-east | per-second | on-demand | Rent → |
| Lambda Labs | $2.21/hr | 1x A100 SXM | us-west-2 | per-minute | on-demand | Rent → |
| Vast.ai | $2.29/hr | 1x H100 SXM | us-east | per-second | on-demand | Rent → |
GPU Requirements
FAQ
What GPU do I need for LLM inference 70B?
Requires at least 80GB VRAM. Recommended: 160GB+. Ideal: NVIDIA H100 80GB SXM, NVIDIA H100 80GB PCIe, NVIDIA H200 141GB SXM.
What is the cheapest GPU for LLM inference 70B?
NVIDIA A100 80GB SXM at $1.15/hr on Paperspace.
How much does LLM inference 70B cost per hour?
From $1.15/hr. Runtime: tokens-per-second.