Quantization format
Q5_K_L
Q5_K_L has a nominal rate of — bits per weight. We have no measured files carrying this label yet.
From the file· 527 files measured
Nominal bpw
—
from the block layout
Measured average
5.964
527 files
Range
4.279–8.803
varies by architecture
File sizes
0.10 GiB+
up to 270.65 GiB
Real files
one per model, most downloaded first
| Model | Size● | Effective bpw● | vs nominal● |
|---|---|---|---|
| Qwen3.6-35B-A3B | 24.43 GiB | 5.836 | — |
| gemma-4-26B-A4B-it | 18.16 GiB | 5.876 | — |
| gemma-4-12B-it | 8.40 GiB | 6.033 | — |
| Qwen3.5-4B | 3.35 GiB | 6.176 | — |
| gemma-4-31B-it | 21.37 GiB | 5.871 | — |
| Llama-3.2-1B-Instruct | 0.91 GiB | 6.312 | — |
| gemma-4-E2B-it | 4.03 GiB | 6.753 | — |
| Qwen3-8B | 5.81 GiB | 6.090 | — |
| Qwen3.5-0.8B | 0.66 GiB | 6.482 | — |
| Llama-3.1-8B-Instruct | 5.64 GiB | 6.034 | — |
| Ornith-1.0-35B | 23.59 GiB | 5.847 | — |
| Llama-3.2-3B-Instruct | 2.25 GiB | 6.020 | — |
| ThinkingCap-Qwen3.6-27B | 20.06 GiB | 6.298 | — |
| Qwythos-9B-v2 | 7.09 GiB | 6.313 | — |
| Qwen3.5-122B-A10B | 84.66 GiB | 5.814 | — |
| Qwen3-Coder-Next | 53.25 GiB | 5.741 | — |
| Qwen2.5-32B-Instruct | 22.11 GiB | 5.797 | — |
| Qwen3.5-35B-A3B | 24.43 GiB | 5.836 | — |
| Qwen2.5-Coder-7B-Instruct | 5.38 GiB | 6.073 | — |
| Qwen2.5-7B-Instruct | 5.38 GiB | 6.073 | — |
| Qwen3-0.6B | 0.60 GiB | 6.891 | — |
| Qwen3-VL-2B-Instruct | 1.24 GiB | 5.013 | — |
| Jan-v3-4B-base-instruct | 3.16 GiB | 6.160 | — |
| Qwen2.5-1.5B-Instruct | 1.10 GiB | 6.123 | — |
| Qwen2.5-VL-7B-Instruct | 5.38 GiB | 5.577 | — |
●From the filewhat these mean