DeepSeek R1 Distill Llama 8B — VRAM Requirements
Written by Jakub Rusinowski · Last updated July 21, 2026
How much GPU VRAM you need to run DeepSeek R1 DeepSeek R1 Distill Llama 8B by DeepSeek locally, a 8B-parameter model. Figures are quantized weights + KV cache + framework overhead from a memory-bandwidth roofline model.
DeepSeek R1 Distill Llama 8B needs about 7 GB VRAM at Q4_K_M.
VRAM by Quantization
| Quant | Bits/weight | Weights | Total VRAM |
|---|---|---|---|
| Q2_K | 2.63 | 2.6 GB | 4.5 GB |
| Q3_K_M | 3.41 | 3.4 GB | 5.3 GB |
| Q4_K_M | 4.83 | 4.8 GB | 6.7 GB |
| Q5_K_M | 5.67 | 5.7 GB | 7.5 GB |
| Q6_K | 6.56 | 6.6 GB | 8.4 GB |
| Q8_0 | 8.50 | 8.5 GB | 10.4 GB |
| F16 | 16.00 | 16.0 GB | 17.9 GB |
Switch quantization in the interactive calculator, or see the full DeepSeek R1 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/deepseek-r1-distill-8b)Estimates only — actual VRAM varies with context length, batch size, runtime and KV-cache settings.