Mistral AI · 119B (6.5B active) · 専家混合（MoE）
Sparse Mistral Small 4 — 6.5B active, strong local all-rounder 看看你的 GPU 或 Mac 跑不跑得動 Mistral Small 4 119B——最低 66.5 GB，建議 110.8 GB。
2026-03256K context
専家混合（MoE）
専家総数： 128
啟用専家： 4
啟用参数： 6.5B
| 量化 | 位元 | VRAM | 品質 | 状態 |
|---|---|---|---|---|
| Q2_K | 2 | 38.6 GB | low | — |
| Q3_K_M | 3 | 53.8 GB | moderate | — |
| Q4_K_M | 4 | 61.5 GB | good | — |
| Q5_K_M | 5 | 76.7 GB | good | — |
| Q6_K | 6 | 91.9 GB | excellent | — |
| Q8_0 | 8 | 122.4 GB | excellent | — |
| F16 | 16 | 244.3 GB | lossless | — |
Can I run Mistral Small 4 119B locally?
- Can I run Mistral Small 4 119B locally?
- Mistral Small 4 119B needs about 66.5 GB of memory at a minimum and 110.8 GB recommended. Open this page to grade it against your GPU or Mac, then run it with runai, Ollama or LM Studio.
- How much VRAM does Mistral Small 4 119B need?
- At Q4_K_M, Mistral Small 4 119B uses about 61.5 GB of VRAM. Higher quants need more memory; lower quants fit tighter cards with a quality tradeoff.