Phi-4 Mini (3.8B) — VRAM Requirements
Autor: Jakub Rusinowski · Ostatnia aktualizacja: 21 lipca 2026
How much GPU VRAM you need to run Phi-4 Mini Phi-4 Mini (3.8B) by Microsoft locally, a 3.8B-parameter model. Figures are quantized weights + KV cache + framework overhead from a memory-bandwidth roofline model.
Phi-4 Mini (3.8B) needs about 3 GB VRAM at Q4_K_M.
VRAM by Quantization
| Quant | Bits/weight | Weights | Total VRAM |
|---|---|---|---|
| Q2_K | 2.63 | 1.2 GB | 2.0 GB |
| Q3_K_M | 3.41 | 1.6 GB | 2.4 GB |
| Q4_K_M | 4.83 | 2.3 GB | 3.1 GB |
| Q5_K_M | 5.67 | 2.7 GB | 3.5 GB |
| Q6_K | 6.56 | 3.1 GB | 3.9 GB |
| Q8_0 | 8.50 | 4.0 GB | 4.8 GB |
| F16 | 16.00 | 7.6 GB | 8.4 GB |
Switch quantization in the interactive calculator, or see the full Phi-4 Mini 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/phi-4-mini-3.8b)Estimates only — actual VRAM varies with context length, batch size, runtime and KV-cache settings.