Quantization format

I1_Q5_K_M

I1_Q5_K_M has a nominal rate of — bits per weight. We have no measured files carrying this label yet.

From the file· 1018 files measured

Nominal bpw

—

from the block layout

Measured average

5.691

1018 files

Range

1.389–12.090

varies by architecture

File sizes

0.12 GiB+

up to 155.36 GiB

Real files

one per model, most downloaded first

Model	Size●	Effective bpw●	vs nominal●
Qwopus3.6-35B-A3B-v1	23.03 GiB	5.503	—
Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-MTP	18.19 GiB	5.625	—
Qwen3-VL-8B-Instruct-abliterated-v1	5.45 GiB	5.339	—
Qwythos-9B-v2	6.19 GiB	5.505	—
gemma-4-26B-A4B-it-ultra-uncensored-heretic	17.82 GiB	5.931	—
Sugoi-14B-Ultra-HF	9.79 GiB	5.692	—
gemma-4-12b-heretic-abliterated	7.96 GiB	5.717	—
FastContext-1.0-4B-SFT	2.69 GiB	5.747	—
Qwen3.6-27B-uncensored-heretic-v2-Native-MTP-Preserved	18.19 GiB	5.713	—
gemma-4-31B-it-The-DECKARD-HERETIC-UNCENSORED-Thinking	20.35 GiB	5.588	—
Qwen3.6-35B-A3B-Claude-4.6-Opus-Reasoning-Distilled	23.03 GiB	5.503	—
Llama3.3-8B-Instruct-Thinking-Heretic-Uncensored-Claude-4.5-Opus-High-Reasoning	5.34 GiB	5.711	—
Huihui-ThinkingCap-Qwen3.6-27B-abliterated	18.19 GiB	5.713	—
Mistral-7B-Instruct-v0.2	4.78 GiB	5.669	—
Huihui-Qwen3-Coder-Next-abliterated	52.89 GiB	5.702	—
DeepSeek-R1-Distill-Llama-70B-abliterated	46.52 GiB	5.664	—
gemma-4-26B-A4B-it-heretic	17.82 GiB	5.931	—
gemma-4-12B-agentic-fable5-composer2.5-v2-3.5x-tau2-uncensored-heretic	7.96 GiB	5.717	—
Qwen3.5-122B-A10B-heretic-v2	80.98 GiB	5.697	—
Llama-3.3-70B-Instruct-abliterated	46.52 GiB	5.664	—
Qwable-9B-Claude-Fable-5	6.02 GiB	5.499	—
EVA-Qwen2.5-14B-v0.2	9.79 GiB	5.692	—
Prosopon-31B	21.25 GiB	5.585	—
Dirty-Muse-Writer-v01-Uncensored-Erotica-NSFW	6.19 GiB	5.754	—
Qwen3.6-34B-80L-Fable-5-Heretic	22.22 GiB	5.708	—

●From the filewhat these mean