Get the app
← back to catalog · registered 2026-08-22 13:56

larue316/Qwen3.6-12B-IQ-Ultra-Heretic-Uncensored-Thinking-V2-Hightop-GGUF

larue316 Qwen 12B GGUF second-order 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/larue316%2FQwen3.6-12B-IQ-Ultra-Heretic-Uncensored-Thinking-V2-Hightop-GGUF"
Response includes
  • classification m3
  • files 18
  • hub_downloads_all_time 12,045
  • author_summary 1 models
  • readme_text full
Try next
GET/models?family=qwen&method=m3
Sibling qwen models · same method
 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
12K
888 last 30d - cooling
Likes
1
Model age
4mo ago
created 2026-06-05
Downloads over time
Now12.4K→from3.7K↑232%
3.3K6.6K9.9K13.2K3.7K on Jun 1012.4K on Oct 11JunJulAugSepOct
Jun 10 → Oct 11 · 59 snapshots · spans 123 days

Genealogy 0 direct forks

Full fork graph →

This model's place in the market. Above: what it was derived from. Below: the tree of everything derived from it.

DavidAU/Qwen3.6-12B-IQ-Ultra-Heretic-Uncensored-Thinking-V2-Hightop↓ 198 this lineage
larue316/Qwen3.6-12B-IQ-Ultra-Heretic-Uncensored-Thinking-V2-Hightop-GGUF↓ 888

Metadata

License
apache-2.0
Languages
en zh
Quantizations
IQ3 IQ4 Q2_K Q3_K Q4 Q4_K Q5 Q5_K Q6_K Q8_0
Tags
gguf qwen uncensored thinking qwen3.5 heretic text-generation en zh base_model:DavidAU/Qwen3.6-12B-IQ-Ultra-Heretic-Uncensored-Thinking-V2-Hightop base_model:quantized:DavidAU/Qwen3.6-12B-IQ-Ultra-Heretic-Uncensored-Thinking-V2-Hightop license:apache-2.0

Related

Same base model
KevinJK51/Qwen3.6-12B-IQ-Ultra-Heretic-Uncensored-Thinking-V2-Hightop-GGUF 13K hellork/Qwen3.6-12B-IQ-Ultra-Heretic-Uncensored-Thinking-V2-Hightop-Q4_K_M-GGUF 922 hellork/Qwen3.6-12B-IQ-Ultra-Heretic-Uncensored-Thinking-V2-Hightop-IQ3_M-GGUF 232
Total size
108 GB
Files
18
Quantizations
11
Registered
2026-08-22 13:56
Last updated on HF
2026-06-05 01:54
Source huggingface.co/larue316/Qwen3.6-12B-IQ-Ultra-Heretic-Uncensored-Thinking-V2-Hightop-GGUF ↗ Base model DavidAU/Qwen3.6-12B-IQ-Ultra-Heretic-Uncensored-Thinking-V2-Hightop →

Files by quantization

Q8_0 1 file 11.6 GB
Qwen3.6-12B-IQ-Q8_0.gguf 11.6 GB dc5ba590 download
Q6_K 1 file 8.93 GB
Qwen3.6-12B-IQ-Q6_K.gguf 8.93 GB 0e1b7e4a download
Q5_K 2 files 15.5 GB
Qwen3.6-12B-IQ-Q5_K_M.gguf 7.84 GB a3888306 download
Qwen3.6-12B-IQ-Q5_K_S.gguf 7.65 GB c3b37f56 download
Q5 1 file 7.65 GB
Qwen3.6-12B-IQ-Q5_0.gguf 7.65 GB b5dcc68c download
Q4_K 2 files 13.3 GB
Qwen3.6-12B-IQ-Q4_K_M.gguf 6.82 GB b01a3594 download
Qwen3.6-12B-IQ-Q4_K_S.gguf 6.49 GB d7a0be05 download
IQ4 2 files 12.9 GB
Qwen3.6-12B-IQ-IQ4_NL.gguf 6.58 GB 06ff150b download
Qwen3.6-12B-IQ-IQ4_XS.gguf 6.31 GB 2d8e1d3d download
Q4 1 file 6.43 GB
Qwen3.6-12B-IQ-Q4_0.gguf 6.43 GB 0fb27550 download
Q3_K 3 files 16.7 GB
Qwen3.6-12B-IQ-Q3_K_L.gguf 5.94 GB 57692a4d download
Qwen3.6-12B-IQ-Q3_K_M.gguf 5.58 GB 2adfc698 download
Qwen3.6-12B-IQ-Q3_K_S.gguf 5.15 GB cb794d38 download
IQ3 2 files 10.6 GB
Qwen3.6-12B-IQ-IQ3_M.gguf 5.33 GB 4394f03e download
Qwen3.6-12B-IQ-IQ3_S.gguf 5.27 GB e60391a1 download
Q2_K 1 file 4.60 GB
Qwen3.6-12B-IQ-Q2_K.gguf 4.60 GB b16853d5 download
Auxiliary files 2 files 3.89 KB
.gitattributes 2.46 KB c0532e15 download
README.md 1.44 KB 7af504aa download

README current version from Hugging Face

license: apache-2.0
base_model: DavidAU/Qwen3.6-12B-IQ-Ultra-Heretic-Uncensored-Thinking-V2-Hightop
language:

  • en
  • zh
    tags:
  • qwen
  • gguf
  • uncensored
  • thinking
  • qwen3.5
  • heretic
  • text-generation

Qwen3.6 12B IQ Ultra Heretic Uncensored Thinking V2 Hightop - GGUF

GGUF quantizations of DavidAU/Qwen3.6-12B-IQ-Ultra-Heretic-Uncensored-Thinking-V2-Hightop.

Converted and quantized using llama.cpp b9192.

Available Quants

Quant Size Quality
Q8_0 11.57 GB Near-perfect
Q6_K 8.93 GB Excellent
Q5_K_M 7.84 GB Very good
Q5_K_S 7.65 GB Very good
Q5_0 7.65 GB Good
Q4_K_M 6.82 GB Best balance
IQ4_NL 6.58 GB Very good (IQ)
Q4_K_S 6.49 GB Good
Q4_0 6.43 GB Good
IQ4_XS 6.31 GB Good (IQ)
Q3_K_L 5.94 GB Acceptable
Q3_K_M 5.58 GB Acceptable
IQ3_M 5.33 GB Acceptable (IQ)
IQ3_S 5.27 GB Acceptable (IQ)
Q3_K_S 5.15 GB Fair
Q2_K 4.60 GB Minimal usable

Original Model

DavidAU/Qwen3.6-12B-IQ-Ultra-Heretic-Uncensored-Thinking-V2-Hightop

Usage

Use with LM Studio, llama.cpp, Ollama, or any GGUF-compatible inference engine.

README history 1 version

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

  1. 2026-06-05Duplicate from KevinJK51/Qwen3.6-12B-IQ-Ultra-Heretic-Uncensored-Thinking-V2-...959eea51.4 KB
    Loading...
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