DeepSeek V4-Flash — VRAM Requirements
作者： Jakub Rusinowski · 最后更新： 2026年7月21日
How much GPU VRAM you need to run DeepSeek V4 DeepSeek V4-Flash by DeepSeek locally, a 284B-parameter model. Figures are quantized weights + KV cache + framework overhead from a memory-bandwidth roofline model.
DeepSeek V4-Flash needs about 172 GB VRAM at Q4_K_M.
VRAM by Quantization
| Quant | Bits/weight | Weights | Total VRAM |
|---|---|---|---|
| Q2_K | 2.63 | 93.4 GB | 94.2 GB |
| Q3_K_M | 3.41 | 121.1 GB | 121.9 GB |
| Q4_K_M | 4.83 | 171.5 GB | 172.3 GB |
| Q5_K_M | 5.67 | 201.3 GB | 202.1 GB |
| Q6_K | 6.56 | 232.9 GB | 233.7 GB |
| Q8_0 | 8.50 | 301.8 GB | 302.6 GB |
| F16 | 16.00 | 568.0 GB | 568.8 GB |
Switch quantization in the interactive calculator, or see the full DeepSeek V4 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-v4-flash)Estimates only — actual VRAM varies with context length, batch size, runtime and KV-cache settings.