AI Training Calculator
Estimate VRAM usage and find the best GPU configuration for your model — inference or fine-tuning.
VRAM Breakdown
21.2GB
- Model Weights16.0 GB75%
- KV Cache0.5 GB3%
- Activations3.2 GB15%
- Framework Overhead1.5 GB7%
GPU Requirements
View Pricing →| GPU | VRAM / Card | Cards Needed | Total VRAM | Utilization |
|---|---|---|---|---|
| A100 40GBRecommended | 40 GB | 1x | 40 GB | 53% |
| B200 | 192 GB | 1x | 192 GB | 11% |
| H200 | 141 GB | 1x | 141 GB | 15% |
| H100 NVL | 94 GB | 1x | 94 GB | 23% |
| A100 80GB | 80 GB | 1x | 80 GB | 27% |
| H100 SXM5 | 80 GB | 1x | 80 GB | 27% |
| L40S | 48 GB | 1x | 48 GB | 44% |
| RTX 4090 | 24 GB | 1x | 24 GB | 89% |