Zhipu AI · 754B (40B active) · 專家混合（MoE）
Improved agentic coding — SOTA SWE-bench Pro, long-horizon tasks 看看你的 GPU 或 Mac 跑不跑得動 GLM-5.1——最低 421.3 GB，建議 702.2 GB。
2026-04128K context
專家混合（MoE）
專家總数： 256
啟用專家： 8
啟用参数： 40.0B
| 量化 | 位元 | VRAM | 品質 | 状態 |
|---|---|---|---|---|
| Q2_K | 2 | 241.9 GB | low | — |
| Q3_K_M | 3 | 338.4 GB | moderate | — |
| Q4_K_M | 4 | 386.7 GB | good | — |
| Q5_K_M | 5 | 483.3 GB | good | — |
| Q6_K | 6 | 579.8 GB | excellent | — |
| Q8_0 | 8 | 772.9 GB | excellent | — |
| F16 | 16 | 1545.4 GB | lossless | — |
Can I run GLM-5.1 locally?
- Can I run GLM-5.1 locally?
- GLM-5.1 needs about 421.3 GB of memory at a minimum and 702.2 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 GLM-5.1 need?
- At Q4_K_M, GLM-5.1 uses about 386.7 GB of VRAM. Higher quants need more memory; lower quants fit tighter cards with a quality tradeoff.