Alibaba · 27B (14B active) · 専家混合（MoE）
Flagship open Wan 2.2 — 14B-active MoE for photoreal text-to-video 看看你的 GPU 或 Mac 跑不跑得動 Wan 2.2 T2V A14B——最低 15.1 GB，建議 25.1 GB。
2025-074K context
専家混合（MoE）
専家総数： 2
啟用専家： 1
啟用参数： 14.0B
| 量化 | 位元 | VRAM | 品質 | 状態 |
|---|---|---|---|---|
| Q2_K | 2 | 9.1 GB | low | — |
| Q3_K_M | 3 | 12.6 GB | moderate | — |
| Q4_K_M | 4 | 14.3 GB | good | — |
| Q5_K_M | 5 | 17.8 GB | good | — |
| Q6_K | 6 | 21.2 GB | excellent | — |
| Q8_0 | 8 | 28.2 GB | excellent | — |
| F16 | 16 | 55.8 GB | lossless | — |
Can I run Wan 2.2 T2V A14B locally?
- Can I run Wan 2.2 T2V A14B locally?
- Wan 2.2 T2V A14B needs about 15.1 GB of memory at a minimum and 25.1 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 Wan 2.2 T2V A14B need?
- At Q4_K_M, Wan 2.2 T2V A14B uses about 14.3 GB of VRAM. Higher quants need more memory; lower quants fit tighter cards with a quality tradeoff.