Command R+ (104B) — VRAM Requirements
作者： Jakub Rusinowski · 最后更新： 2026年7月21日
How much GPU VRAM you need to run Command R Family Command R+ (104B) by Cohere locally, a 104B-parameter model. Figures are quantized weights + KV cache + framework overhead from a memory-bandwidth roofline model.
Command R+ (104B) needs about 64 GB VRAM at Q4_K_M.
VRAM by Quantization
| Quant | Bits/weight | Weights | Total VRAM |
|---|---|---|---|
| Q2_K | 2.63 | 34.2 GB | 35.0 GB |
| Q3_K_M | 3.41 | 44.3 GB | 45.1 GB |
| Q4_K_M | 4.83 | 62.8 GB | 63.6 GB |
| Q5_K_M | 5.67 | 73.7 GB | 74.5 GB |
| Q6_K | 6.56 | 85.3 GB | 86.1 GB |
| Q8_0 | 8.50 | 110.5 GB | 111.3 GB |
| F16 | 16.00 | 208.0 GB | 208.8 GB |
Switch quantization in the interactive calculator, or see the full Command R Family 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/command-r-plus)Estimates only — actual VRAM varies with context length, batch size, runtime and KV-cache settings.