24GB VRAM GPU Cloud Rental for AI & LLMs
12 GPU models with approximately 24GB VRAM, prices from $0.021/hr. Best for: 13B model inference, SDXL, lightweight training.
Last updated May 26, 2026 · Data refreshed every 6 hours
GPU Models
12
Cheapest
$0.021/hr
Best $/GB
$0.001/GB-hr
Total Instances
754
Will this VRAM tier run my model?
24GB VRAM is best for 13B model inference, SDXL, lightweight training. For LLM inference, leave headroom for quantization format, batch size, and KV cache: long context windows can add meaningful VRAM pressure even when the base model fits.
All 24GB VRAM GPUs
| GPU Model | Providers | Instances | From / GB | From |
|---|---|---|---|---|
| RTX3090 | 2 | 50 | $0.001/GB-hr | $0.021/hr |
| RTXPRO4000 | 1 | 2 | $0.004/GB-hr | $0.107/hr |
| RTX4090 | 3 | 96 | $0.005/GB-hr | $0.131/hr |
| A5000 | 1 | 12 | $0.006/GB-hr | $0.140/hr |
| RTX3090Ti | 1 | 8 | $0.006/GB-hr | $0.140/hr |
| L4 | 9 | 374 | $0.008/GB-hr | $0.188/hr |
| RTX 3090 | 5 | 6 | $0.010/GB-hr | $0.240/hr |
| A30 | 1 | 4 | $0.015/GB-hr | $0.350/hr |
| A10 | 5 | 172 | $0.015/GB-hr | $0.360/hr |
| RTX 4090 | 11 | 18 | $0.018/GB-hr | $0.440/hr |
| RTX A5000 | 3 | 6 | $0.018/GB-hr | $0.440/hr |
| A10G | 4 | 6 | $0.029/GB-hr | $0.710/hr |
Other VRAM Tiers