Quantization format

I1_Q5_K_S

I1_Q5_K_S has a nominal rate of — bits per weight. We have no measured files carrying this label yet.

From the file· 1016 files measured

Nominal bpw

—

from the block layout

Measured average

5.536

1016 files

Range

1.333–11.804

varies by architecture

File sizes

0.11 GiB+

up to 150.76 GiB

Real files

one per model, most downloaded first

Model	Size●	Effective bpw●	vs nominal●
Qwopus3.6-35B-A3B-v1	22.33 GiB	5.336	—
Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-MTP	17.67 GiB	5.463	—
Qwen3-VL-8B-Instruct-abliterated-v1	5.33 GiB	5.220	—
Qwythos-9B-v2	6.03 GiB	5.364	—
gemma-4-26B-A4B-it-ultra-uncensored-heretic	16.75 GiB	5.576	—
Sugoi-14B-Ultra-HF	9.56 GiB	5.561	—
gemma-4-12b-heretic-abliterated	7.77 GiB	5.578	—
FastContext-1.0-4B-SFT	2.63 GiB	5.616	—
Qwen3.6-27B-uncensored-heretic-v2-Native-MTP-Preserved	17.67 GiB	5.548	—
gemma-4-31B-it-The-DECKARD-HERETIC-UNCENSORED-Thinking	19.85 GiB	5.452	—
Qwen3.6-35B-A3B-Claude-4.6-Opus-Reasoning-Distilled	22.33 GiB	5.336	—
Llama3.3-8B-Instruct-Thinking-Heretic-Uncensored-Claude-4.5-Opus-High-Reasoning	5.21 GiB	5.578	—
Huihui-ThinkingCap-Qwen3.6-27B-abliterated	17.67 GiB	5.548	—
Mistral-7B-Instruct-v0.2	4.65 GiB	5.521	—
Huihui-Qwen3-Coder-Next-abliterated	51.22 GiB	5.522	—
DeepSeek-R1-Distill-Llama-70B-abliterated	45.32 GiB	5.517	—
gemma-4-26B-A4B-it-heretic	16.75 GiB	5.576	—
gemma-4-12B-agentic-fable5-composer2.5-v2-3.5x-tau2-uncensored-heretic	7.77 GiB	5.578	—
Qwen3.5-122B-A10B-heretic-v2	78.41 GiB	5.516	—
Llama-3.3-70B-Instruct-abliterated	45.32 GiB	5.517	—
Qwable-9B-Claude-Fable-5	5.87 GiB	5.361	—
EVA-Qwen2.5-14B-v0.2	9.56 GiB	5.561	—
Prosopon-31B	20.75 GiB	5.454	—
Dirty-Muse-Writer-v01-Uncensored-Erotica-NSFW	6.04 GiB	5.612	—
Qwen3.6-34B-80L-Fable-5-Heretic	21.57 GiB	5.540	—

●From the filewhat these mean