Quantization format
UD_Q3_K_M
UD_Q3_K_M has a nominal rate of — bits per weight. We have no measured files carrying this label yet.
From the file· 37 files measured
Nominal bpw
—
from the block layout
Measured average
4.412
37 files
Range
3.369–11.747
varies by architecture
File sizes
3.67 GiB+
up to 431.78 GiB
Real files
one per model, most downloaded first
| Model | Size● | Effective bpw● | vs nominal● |
|---|---|---|---|
| Qwen3.6-35B-A3B | 15.46 GiB | 3.694 | — |
| gemma-4-26B-A4B-it | 11.85 GiB | 3.836 | — |
| DeepSeek-V4-Flash | 120.44 GiB | 3.556 | — |
| Qwen-AgentWorld-35B-A3B | 15.53 GiB | 3.850 | — |
| LTX-2.3 | 12.26 GiB | 11.455 | — |
| GLM-5.2 | 319.20 GiB | 3.640 | — |
| Ornith-1.0-35B | 15.53 GiB | 3.850 | — |
| Kimi-K2.7-Code | 431.78 GiB | 3.504 | — |
| Qwen3.5-122B-A10B | 54.20 GiB | 3.722 | — |
| Qwen3-Coder-Next | 33.47 GiB | 3.608 | — |
| LTX-2 | 9.39 GiB | 4.271 | — |
| Laguna-S-2.1 | 50.31 GiB | 3.676 | — |
| Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 | 18.19 GiB | 4.734 | — |
| Step-3.7-Flash | 83.13 GiB | 3.546 | — |
| LFM2.5-8B-A1B | 3.67 GiB | 3.722 | — |
| Ornith-1.0-397B | 167.62 GiB | 3.629 | — |
| Qwen3.5-397B-A17B | 165.23 GiB | 3.518 | — |
| North-Mini-Code-1.0 | 13.24 GiB | 3.730 | — |
| MiniMax-M2.7 | 94.29 GiB | 3.542 | — |
| MiniMax-M3 | 181.22 GiB | 3.645 | — |
| NVIDIA-Nemotron-3-Super-120B-A12B-BF16 | 57.47 GiB | 3.994 | — |
| Mistral-Small-4-119B-2603 | 50.64 GiB | 3.643 | — |
| Huihui-gemma-4-26B-A4B-it-abliterated | 11.67 GiB | 3.775 | — |
| MiMo-V2.5 | 130.41 GiB | 3.605 | — |
| GLM-5.1 | 315.19 GiB | 3.591 | — |
●From the filewhat these mean