Qwen3-Coder 80B-A3B (MoE) — VRAM Requirements
Autor: Jakub Rusinowski · Ostatnia aktualizacja: 21 lipca 2026
How much GPU VRAM you need to run Qwen3-Coder Qwen3-Coder 80B-A3B (MoE) by Alibaba Cloud locally, a 80B-parameter model. Figures are quantized weights + KV cache + framework overhead from a memory-bandwidth roofline model.
Qwen3-Coder 80B-A3B (MoE) needs about 49 GB VRAM at Q4_K_M.
VRAM by Quantization
| Quant | Bits/weight | Weights | Total VRAM |
|---|---|---|---|
| Q2_K | 2.63 | 26.3 GB | 27.1 GB |
| Q3_K_M | 3.41 | 34.1 GB | 34.9 GB |
| Q4_K_M | 4.83 | 48.3 GB | 49.1 GB |
| Q5_K_M | 5.67 | 56.7 GB | 57.5 GB |
| Q6_K | 6.56 | 65.6 GB | 66.4 GB |
| Q8_0 | 8.50 | 85.0 GB | 85.8 GB |
| F16 | 16.00 | 160.0 GB | 160.8 GB |
Switch quantization in the interactive calculator, or see the full Qwen3-Coder 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/qwen3-coder-80b-a3b)Estimates only — actual VRAM varies with context length, batch size, runtime and KV-cache settings.