Quantization format

I1_Q4_0

I1_Q4_0 has a nominal rate of — bits per weight. We have no measured files carrying this label yet.

From the file· 1033 files measured

Nominal bpw

—

from the block layout

Measured average

4.625

1033 files

Range

1.131–9.934

varies by architecture

File sizes

0.10 GiB+

up to 162.70 GiB

Real files

one per model, most downloaded first

Model	Size●	Effective bpw●	vs nominal●
Qwopus3.6-35B-A3B-v1	18.44 GiB	4.406	—
Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-MTP	14.68 GiB	4.538	—
Qwen3-VL-8B-Instruct-abliterated-v1	4.46 GiB	4.368	—
Qwythos-9B-v2	5.09 GiB	4.527	—
gemma-4-26B-A4B-it-ultra-uncensored-heretic	13.49 GiB	4.491	—
Sugoi-14B-Ultra-HF	7.96 GiB	4.628	—
gemma-4-12b-heretic-abliterated	6.52 GiB	4.681	—
FastContext-1.0-4B-SFT	2.21 GiB	4.725	—
Qwen3.6-27B-uncensored-heretic-v2-Native-MTP-Preserved	14.68 GiB	4.609	—
gemma-4-31B-it-The-DECKARD-HERETIC-UNCENSORED-Thinking	16.49 GiB	4.528	—
Qwen3.6-35B-A3B-Claude-4.6-Opus-Reasoning-Distilled	18.44 GiB	4.406	—
Llama3.3-8B-Instruct-Thinking-Heretic-Uncensored-Claude-4.5-Opus-High-Reasoning	4.35 GiB	4.658	—
Huihui-ThinkingCap-Qwen3.6-27B-abliterated	14.68 GiB	4.609	—
Mistral-7B-Instruct-v0.2	3.84 GiB	4.555	—
Huihui-Qwen3-Coder-Next-abliterated	42.17 GiB	4.547	—
DeepSeek-R1-Distill-Llama-70B-abliterated	37.36 GiB	4.549	—
gemma-4-26B-A4B-it-heretic	13.49 GiB	4.491	—
gemma-4-12B-agentic-fable5-composer2.5-v2-3.5x-tau2-uncensored-heretic	6.52 GiB	4.681	—
Qwen3.5-122B-A10B-heretic-v2	64.57 GiB	4.543	—
Llama-3.3-70B-Instruct-abliterated	37.36 GiB	4.549	—
Qwable-9B-Claude-Fable-5	4.96 GiB	4.528	—
EVA-Qwen2.5-14B-v0.2	7.96 GiB	4.628	—
Prosopon-31B	17.22 GiB	4.527	—
Dirty-Muse-Writer-v01-Uncensored-Erotica-NSFW	5.08 GiB	4.726	—
Qwen3.6-34B-80L-Fable-5-Heretic	17.88 GiB	4.592	—

●From the filewhat these mean