Live Floor
$0.0689
Vast · FR
The NVIDIA GeForce RTX 3070 is built on the Ampere architecture with 8 GB of GDDR6 VRAM.
Best live route via Vast · FR
Live Floor
$0.0689
Vast · FR
Average Price
$0.1379
Across 22 live offers
Reliability
98.9%
Based on 22 scored hosts
Last Day Delta
-$0.0067
Latest daily low: 2026-07-15
30-Day Price Trend
Available Offers
Sorted by cheapest hourly rate
| Provider | Price/hr | VRAM | GPUs | Reliability | Region | |
|---|---|---|---|---|---|---|
| Vast | $0.0689 | 8 GB | 1x | 99.1% | FR | View Deal |
| Vast | $0.0756 | 8 GB | 1x | 99.9% | CA | View Deal |
| Vast | $0.0778 | 8 GB | 1x | 99.6% | US | View Deal |
| Vast | $0.0822 | 8 GB | 1x | 98.8% | US | View Deal |
| Vast | $0.0889 | 8 GB | 1x | 99.9% | CA | View Deal |
| Vast | $0.0916 | 8 GB | 1x | 99.8% | SE | View Deal |
| Vast | $0.1015 | 8 GB | 1x | 99.9% | CA | View Deal |
| Vast | $0.1077 | 8 GB | 1x | 99.9% | US | View Deal |
| Vast | $0.1081 | 8 GB | 1x | 98.9% | US | View Deal |
| Vast | $0.1196 | 8 GB | 1x | 98.9% | ES | View Deal |
| RunPod | $0.1300 | 8 GB | 1x | 90.0% | -- | View Deal |
| Vast | $0.1356 | 8 GB | 1x | 99.2% | CZ | View Deal |
| Vast | $0.1370 | 8 GB | 1x | 99.9% | NO | View Deal |
| Vast | $0.1422 | 8 GB | 1x | 99.7% | FR | View Deal |
| Vast | $0.1489 | 8 GB | 2x | 99.9% | CA | View Deal |
| Vast | $0.1511 | 8 GB | 2x | 99.6% | US | View Deal |
| Vast | $0.1970 | 8 GB | 1x | 97.0% | ES | View Deal |
| Vast | $0.1996 | 8 GB | 1x | 96.6% | TH | View Deal |
| Vast | $0.2015 | 8 GB | 2x | 99.9% | CA | View Deal |
| Vast | $0.2156 | 8 GB | 2x | 99.7% | CA | View Deal |
| Vast | $0.2170 | 8 GB | 2x | 99.9% | US | View Deal |
| Vast | $0.2369 | 8 GB | 2x | 98.9% | ES | View Deal |
GPU Overview
The NVIDIA GeForce RTX 3070 is built on the Ampere architecture with 8 GB of GDDR6 VRAM. It represents a budget-friendly entry point into local AI development, delivering good value for money for developers and hobbyists getting started with GPU compute.
Its 8 GB of VRAM accommodates Stable Diffusion and creative workflows at moderate settings and batch sizes. The card handles inference on small quantized models and provides a practical environment for learning AI development techniques and prototyping lightweight pipelines.