Kimi K3 2.8T — VRAM Requirements
Autor: Jakub Rusinowski · Ostatnia aktualizacja: 30 lipca 2026
How much GPU VRAM you need to run Kimi K3 Kimi K3 2.8T by Moonshot AI locally, a 2800B-parameter model. Figures are quantized weights + KV cache + framework overhead from a memory-bandwidth roofline model.
Kimi K3 2.8T needs about 1691 GB VRAM at Q4_K_M.
VRAM by Quantization
| Quant | Bits/weight | Weights | Total VRAM |
|---|---|---|---|
| Q2_K | 2.63 | 920.5 GB | 921.3 GB |
| Q3_K_M | 3.41 | 1193.5 GB | 1194.3 GB |
| Q4_K_M | 4.83 | 1690.5 GB | 1691.3 GB |
| Q5_K_M | 5.67 | 1984.5 GB | 1985.3 GB |
| Q6_K | 6.56 | 2296.0 GB | 2296.8 GB |
| Q8_0 | 8.50 | 2975.0 GB | 2975.8 GB |
| F16 | 16.00 | 5600.0 GB | 5600.8 GB |
Switch quantization in the interactive calculator, or see the full Kimi K3 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/kimi-k3-2-8t)Estimates only — actual VRAM varies with context length, batch size, runtime and KV-cache settings.