Quantization format

I1_Q4_K_S

I1_Q4_K_S has a nominal rate of — bits per weight. We have no measured files carrying this label yet.

From the file· 1034 files measured
Nominal bpw
—
from the block layout
Measured average
4.672
1034 files
Range
1.131–9.965
varies by architecture
File sizes
0.10 GiB+
up to 163.33 GiB

Real files

one per model, most downloaded first
ModelSize●Effective bpw●vs nominal●
Qwopus3.6-35B-A3B-v118.52 GiB4.426—
Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-MTP14.74 GiB4.557—
Qwen3-VL-8B-Instruct-abliterated-v14.47 GiB4.382—
Qwythos-9B-v25.11 GiB4.549—
gemma-4-26B-A4B-it-ultra-uncensored-heretic14.40 GiB4.794—
Sugoi-14B-Ultra-HF7.98 GiB4.644—
gemma-4-12b-heretic-abliterated6.54 GiB4.699—
FastContext-1.0-4B-SFT2.22 GiB4.740—
Qwen3.6-27B-uncensored-heretic-v2-Native-MTP-Preserved14.74 GiB4.628—
gemma-4-31B-it-The-DECKARD-HERETIC-UNCENSORED-Thinking16.54 GiB4.544—
Qwen3.6-35B-A3B-Claude-4.6-Opus-Reasoning-Distilled18.52 GiB4.426—
Llama3.3-8B-Instruct-Thinking-Heretic-Uncensored-Claude-4.5-Opus-High-Reasoning4.37 GiB4.675—
Huihui-ThinkingCap-Qwen3.6-27B-abliterated14.74 GiB4.628—
Mistral-7B-Instruct-v0.23.86 GiB4.574—
Huihui-Qwen3-Coder-Next-abliterated42.36 GiB4.567—
DeepSeek-R1-Distill-Llama-70B-abliterated37.58 GiB4.575—
gemma-4-26B-A4B-it-heretic14.40 GiB4.794—
gemma-4-12B-agentic-fable5-composer2.5-v2-3.5x-tau2-uncensored-heretic6.54 GiB4.699—
Qwen3.5-122B-A10B-heretic-v264.86 GiB4.562—
Llama-3.3-70B-Instruct-abliterated37.58 GiB4.575—
Qwable-9B-Claude-Fable-54.98 GiB4.550—
EVA-Qwen2.5-14B-v0.27.98 GiB4.644—
Prosopon-31B17.28 GiB4.542—
Dirty-Muse-Writer-v01-Uncensored-Erotica-NSFW5.10 GiB4.743—
Qwen3.6-34B-80L-Fable-5-Heretic17.95 GiB4.610—
●From the filewhat these mean

Other quantizations

Q2_KQ3_K_SQ3_K_MQ3_K_LQ4_0Q4_K_SQ4_K_MQ5_K_SQ5_K_MQ6_KQ8_0IQ4_XSIQ3_XXSIQ2_XXSF16BF16F32