Quantization format

I1_Q6_K

I1_Q6_K has a nominal rate of — bits per weight. We have no measured files carrying this label yet.

From the file· 990 files measured
Nominal bpw
—
from the block layout
Measured average
6.541
990 files
Range
1.548–13.822
varies by architecture
File sizes
0.13 GiB+
up to 179.76 GiB

Real files

one per model, most downloaded first
ModelSize●Effective bpw●vs nominal●
Qwopus3.6-35B-A3B-v126.56 GiB6.345—
Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-MTP20.89 GiB6.459—
Qwen3-VL-8B-Instruct-abliterated-v16.26 GiB6.137—
Qwythos-9B-v27.04 GiB6.264—
gemma-4-26B-A4B-it-ultra-uncensored-heretic21.08 GiB7.018—
Sugoi-14B-Ultra-HF11.29 GiB6.567—
gemma-4-12b-heretic-abliterated9.11 GiB6.546—
FastContext-1.0-4B-SFT3.08 GiB6.576—
Qwen3.6-27B-uncensored-heretic-v2-Native-MTP-Preserved20.89 GiB6.560—
gemma-4-31B-it-The-DECKARD-HERETIC-UNCENSORED-Thinking23.47 GiB6.447—
Qwen3.6-35B-A3B-Claude-4.6-Opus-Reasoning-Distilled26.56 GiB6.345—
Llama3.3-8B-Instruct-Thinking-Heretic-Uncensored-Claude-4.5-Opus-High-Reasoning6.14 GiB6.571—
Huihui-ThinkingCap-Qwen3.6-27B-abliterated20.89 GiB6.560—
Mistral-7B-Instruct-v0.25.53 GiB6.564—
Huihui-Qwen3-Coder-Next-abliterated61.03 GiB6.580—
gemma-4-26B-A4B-it-heretic21.08 GiB7.018—
gemma-4-12B-agentic-fable5-composer2.5-v2-3.5x-tau2-uncensored-heretic9.11 GiB6.546—
Qwen3.5-122B-A10B-heretic-v293.42 GiB6.572—
Qwable-9B-Claude-Fable-56.85 GiB6.257—
EVA-Qwen2.5-14B-v0.211.29 GiB6.567—
Prosopon-31B24.55 GiB6.452—
Dirty-Muse-Writer-v01-Uncensored-Erotica-NSFW7.07 GiB6.569—
Qwen3.6-34B-80L-Fable-5-Heretic25.54 GiB6.561—
Gemma-4-26B-A4B-StyleTune-V221.65 GiB7.005—
Qwen2.5-VL-7B-Instruct-abliterated5.82 GiB6.034—
●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