Llama 4 Scout 17B — VRAM Requirements
Written by Jakub Rusinowski · Last updated July 21, 2026
How much GPU VRAM you need to run Llama 4 Llama 4 Scout 17B by Meta locally, a 109B-parameter model. Figures are quantized weights + KV cache + framework overhead from a memory-bandwidth roofline model.
Llama 4 Scout 17B needs about 67 GB VRAM at Q4_K_M.
VRAM by Quantization
| Quant | Bits/weight | Weights | Total VRAM |
|---|---|---|---|
| Q2_K | 2.63 | 35.8 GB | 36.6 GB |
| Q3_K_M | 3.41 | 46.5 GB | 47.3 GB |
| Q4_K_M | 4.83 | 65.8 GB | 66.6 GB |
| Q5_K_M | 5.67 | 77.3 GB | 78.1 GB |
| Q6_K | 6.56 | 89.4 GB | 90.2 GB |
| Q8_0 | 8.50 | 115.8 GB | 116.6 GB |
| F16 | 16.00 | 218.0 GB | 218.8 GB |
Switch quantization in the interactive calculator, or see the full Llama 4 model page.
Add this badge to your model card
Model creators: paste this into your Hugging Face model card README to link readers straight to this VRAM breakdown.
[](https://llmconfigurator.com/en/vram-calculator/llama-4-scout-17b)Estimates only — actual VRAM varies with context length, batch size, runtime and KV-cache settings.