Quantization format

I1_IQ1_S

I1_IQ1_S has a nominal rate of — bits per weight. We have no measured files carrying this label yet.

From the file· 957 files measured

Nominal bpw

—

from the block layout

Measured average

2.068

957 files

Range

1.351–5.114

varies by architecture

File sizes

0.06 GiB+

up to 73.49 GiB

Real files

one per model, most downloaded first

Model	Size●	Effective bpw●	vs nominal●
Qwopus3.6-35B-A3B-v1	6.97 GiB	1.665	—
Qwen3-VL-8B-Instruct-abliterated-v1	1.97 GiB	1.931	—
gemma-4-26B-A4B-it-ultra-uncensored-heretic	7.72 GiB	2.570	—
Sugoi-14B-Ultra-HF	3.36 GiB	1.954	—
gemma-4-12b-heretic-abliterated	2.78 GiB	1.996	—
FastContext-1.0-4B-SFT	0.98 GiB	2.099	—
gemma-4-31B-it-The-DECKARD-HERETIC-UNCENSORED-Thinking	6.66 GiB	1.831	—
Qwen3.6-35B-A3B-Claude-4.6-Opus-Reasoning-Distilled	6.97 GiB	1.665	—
Llama3.3-8B-Instruct-Thinking-Heretic-Uncensored-Claude-4.5-Opus-High-Reasoning	1.88 GiB	2.012	—
Mistral-7B-Instruct-v0.2	1.50 GiB	1.781	—
Huihui-Qwen3-Coder-Next-abliterated	15.07 GiB	1.625	—
DeepSeek-R1-Distill-Llama-70B-abliterated	14.29 GiB	1.740	—
gemma-4-26B-A4B-it-heretic	7.72 GiB	2.570	—
gemma-4-12B-agentic-fable5-composer2.5-v2-3.5x-tau2-uncensored-heretic	2.78 GiB	1.996	—
Qwen3.5-122B-A10B-heretic-v2	23.13 GiB	1.627	—
Llama-3.3-70B-Instruct-abliterated	14.29 GiB	1.740	—
Qwable-9B-Claude-Fable-5	2.55 GiB	2.332	—
EVA-Qwen2.5-14B-v0.2	3.36 GiB	1.954	—
Prosopon-31B	7.10 GiB	1.865	—
Dirty-Muse-Writer-v01-Uncensored-Erotica-NSFW	2.22 GiB	2.059	—
Gemma-4-26B-A4B-StyleTune-V2	7.95 GiB	2.572	—
Qwen2.5-VL-7B-Instruct-abliterated	1.77 GiB	1.837	—
Qwen3.5-9B-Claude-4.6-Opus-Deckard-V4.2-Uncensored-Heretic-Thinking	2.55 GiB	2.332	—
MythoMax-L2-13b	2.90 GiB	1.914	—
Mistral-7B-Instruct-v0.1	1.50 GiB	1.781	—

●From the filewhat these mean