
The Internet's Cheapest GPU Marketplace
Current pricing insights across all providers and asset types
Tell us what you're building. We'll recommend the hardware and show you the best deals across all providers.

For training 70B+ parameter LLMs, H100's 80GB HBM3 memory and 1,979 TFLOPS FP8 performance are critical. The advanced tensor cores accelerate transformer attention mechanisms by 3x versus A100, while NVLink enables efficient multi-GPU scaling for models that exceed single-GPU capacity.
Model complex infrastructure scenarios. Adjust variables below to forecast your 30-day capital expenditure.
Popular GPUs at best available prices
Compare all available asset configurations side-by-side. Prices are sorted from lowest to highest to help you find the best deals.
Asset Type | Provider | VRAM | Price / Hour | |
|---|---|---|---|---|
| 80 GB | $1.21 /GPU/hr | |||
Vast.aiMarketplace | 80 GB | $1.74 /GPU/hr | ||
| 80 GB | $2.40 /GPU/hr | — | ||
| 80 GB | $2.69 /GPU/hr | |||
| 80 GB | $2.99 /GPU/hr | |||
| 80 GB | $3.90 /GPU/hr | — | ||
| 80 GB | $11.06 /GPU/hr | — | ||
| 192 GB | $1.16 /GPU/hr | |||
| 192 GB | $2.89 /GPU/hr | — | ||
| 192 GB | $4.99 /GPU/hr | |||
| 192 GB | $5.19 /GPU/hr | |||
| 192 GB | $10.00 /GPU/hr | — | ||
| 192 GB | $14.24 /GPU/hr | — | ||
| 80 GB | $2.39 /GPU/hr | |||
| 80 GB | $2.40 /GPU/hr | |||
| 80 GB | $6.98 /GPU/hr | — | ||
| 141 GB | $1.79 /GPU/hr | |||
Vast.aiMarketplace | 141 GB | $2.01 /GPU/hr | ||
| 141 GB | $3.39 /GPU/hr | |||
| 141 GB | $3.99 /GPU/hr | — | ||
| 141 GB | $4.33 /GPU/hr | — | ||
| 141 GB | $10.71 /GPU/hr | — | ||
| 141 GB | $13.78 /GPU/hr | — | ||
| 192 GB | $1.49 /GPU/hr | |||
| 80 GB | $0.32 /GPU/hr | |||
| 80 GB | $1.49 /GPU/hr | |||
| 80 GB | $1.79 /GPU/hr | |||
| 80 GB | $4.10 /GPU/hr | — | ||
| 80 GB | $6.25 /GPU/hr | — | ||
| 80 GB | $1.39 /GPU/hr | |||
| 80 GB | $1.52 /GPU/hr | |||
| 80 GB | $2.40 /GPU/hr | — | ||
| 80 GB | $3.67 /GPU/hr | — | ||
| 48 GB | $0.37 /GPU/hr | |||
| 48 GB | $0.86 /GPU/hr | |||
| 48 GB | $1.67 /GPU/hr | — | ||
| 48 GB | $3.77 /GPU/hr | — | ||
| 96 GB | $0.47 /GPU/hr | |||
| 96 GB | $0.50 /GPU/hr | |||
Vast.aiMarketplace | 96 GB | $0.51 /GPU/hr | ||
Vast.aiMarketplace | 32 GB | $0.36 /GPU/hr | ||
| 32 GB | $0.73 /GPU/hr | |||
| 32 GB | $0.89 /GPU/hr | |||
| 24 GB | $0.55 /GPU/hr | |||
| 24 GB | $0.59 /GPU/hr | |||
| 48 GB | $0.72 /GPU/hr | |||
| 48 GB | $0.99 /GPU/hr | |||
Vast.aiMarketplace | 48 GB | $0.46 /GPU/hr | ||
| 48 GB | $0.64 /GPU/hr | |||
| 48 GB | $0.77 /GPU/hr |
Everything you need to know about renting GPUs for machine learning, AI training, and inference workloads.
Still have questions? Contact us