Alibaba/Qwen · open-weights model
Qwen3.6-35B-A3B
Qwen3.6-35B-A3B is a 35B-parameter model (MoE, 3B active) by Alibaba/Qwen. Its smallest quant (MXFP4_MOE) needs at least 22.8GB of GPU-addressable memory at 8K context; it loads cleanly at a quality quant on 16 of the 50 machines we track.
- Parameters
- 35B total · 3B active
- Layers
- 40
- Max context
- 256K
- Attention
- HYBRID
- License
- apache-2.0
Config from Qwen/Qwen3.6-35B-A3B
Quants and what they need
Measured file sizes; “needs” = weights + KV cache at 8K context + overhead.
| Quant | File size | Needs ≥ |
|---|---|---|
| Q8_0 | 36.9 GB | 38.0 GB |
| UD-Q5_K_XL | 26.6 GB | 27.7 GB |
| UD-Q4_K_XL | 22.4 GB | 23.4 GB |
| MXFP4_MOE | 21.7 GB | 22.8 GB |
What runs Qwen3.6-35B-A3B
Every machine we track, graded at its best clean-fit quant.
Verdicts computed from published specs and measured GGUF sizes — how the math works.