Get the app
← back to catalog · registered 2026-09-12 12:55

mradermacher/Qwen3.8-27B-Uncensored-Heretic-v3-i1-GGUF

mradermacher 27B GGUF 262K ctx
Your rig guess connected
? Why do I need an app?
Reading your rig…

This is a rough estimate. Install the free app - we'll show exact numbers.

Reading real hardware from your app right now. Numbers below are exact.

Below is the per-quantization compatibility for this model. 

curl -H "Authorization: Bearer $ABL_KEY" \
     "https://abliteration.org/api/v1/models/mradermacher%2FQwen3.8-27B-Uncensored-Heretic-v3-i1-GGUF"
Response includes
  • classification m3
  • files 5
  • hub_downloads_all_time 5,082
  • author_summary 3413 models
  • readme_text full
Try next
GET/models?author=mradermacher
All 3413 models by mradermacher
 GET/methods/m3
What "m3" means · taxonomy
10 credits · hourly refresh · ~4 KB payload Get an API key →
Abliteration classifier · v1.0.0
M3
Primary method

Layer-wise ablation

Applied on top of direct removal inherited from the base model.
Confidence
HIGH
Inherited from base model
M1 Direct removal →
Why this label 2 signals
Producer identity confirmed by naming conventions, tags or the model card. This label is very unlikely to change.
  • 'heretic' in model name (Heretic-produced)
  • Heretic uses layer-wise optimization (M3) with underlying direction removal (M1)
How we label → About M3 →
Refusal direction extracted via
Extraction technique

Difference-of-means

FromArditi et al. (2024) ↗
Confidence
HIGH
Why we say so
name contains 'heretic'; Heretic default extraction is difference-of-means (Arditi 2024)
What is a refusal direction? → Read the paper ↗
Downloads · lifetime
5K
1K last 30d - stable
Likes
4
Model age
6w ago
created 2026-09-12
Downloads over time
Now6.3K→from0↑0%
02.3K4.7K7K0 on Sep 126.3K on Oct 28SepOct
Sep 12 → Oct 28 · 47 snapshots · spans 46 days

Variants by this author 2 formats · 4K downloads combined

The same weights this author released in different packaging. Pick the format that matches your runtime.

Metadata

Quantizations
IQ3 Q2_K
Tags
transformers gguf heretic uncensored decensored abliterated ara-lora en base_model:OS-Software/Qwen3.8-27B-Uncensored-Heretic-v3 base_model:quantized:OS-Software/Qwen3.8-27B-Uncensored-Heretic-v3 license:apache-2.0 endpoints_compatible

Related

Other by mradermacher
Ornith-1.5-9B-uncensored-GGUF 1.7M Qwen3-VL-8B-Instruct-abliterated-GGUF 663K Huihui-Qwen3.5-9B-abliterated-GGUF 177K Qwen3.8-Flash-Next-Uncensored-i1-GGUF 159K Huihui-Qwen3.8-Flash-Next-abliterated-i1-GGUF 107K
Total size
21.7 GB
Files
5
Quantizations
3
Registered
2026-09-12 12:55
Last updated on HF
2026-09-12 14:41
Source huggingface.co/mradermacher/Qwen3.8-27B-Uncensored-Heretic-v3-i1-GGUF ↗

Files by quantization

IQ3 1 file 11.7 GB
Qwen3.8-27B-Uncensored-Heretic-v3.i1-IQ3_M.gguf 11.7 GB 5f74f5a1 download
Q2_K 1 file 9.98 GB
Qwen3.8-27B-Uncensored-Heretic-v3.i1-Q2_K.gguf 9.98 GB 9638108b download
Auxiliary files 3 files 13.0 MB
Qwen3.8-27B-Uncensored-Heretic-v3.imatrix.gguf 13.0 MB 33dc7d3e download
.gitattributes 1.73 KB bea5b5b4 download
README.md 491 B 514f088e download

README current version from Hugging Face

README history 3 versions

The author's README evolved over time. Click a version to see its content at that point.

  1. 2026-09-12auto-patch README.md1c624966.8 KB
    Loading...
  2. 2026-09-12auto-patch README.md9cb9bbd5.2 KB
    Loading...
  3. 2026-09-12uploaded from nico1009ce3f491 B
    Loading...

Discussions 1 thread

  1. 2026-09-18MTP missing/not workingopen2 💬#1
    Loading...

Continue on Hugging Face →

Catalog is the map. Apps are the tools.

Run models on your own machine, not in the cloud.

Every model page has an "Open in Abliteration" button that hands the model directly to the first-party desktop client, at the quantization your rig can actually run. No API keys, no subscription, no prompt leakage.

Get the app Browse the catalog
Open in Abliteration