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

ijwfly/Qwen3.6-35B-A3B-uncensored-heretic-mlx-8Bit

ijwfly Qwen 35B MoE multimodal second-order
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/ijwfly%2FQwen3.6-35B-A3B-uncensored-heretic-mlx-8Bit"
Response includes
  • classification m3
  • files 19
  • benchmarks 11 entries
  • hub_downloads_all_time 2,607
  • 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
3K
72 last 30d - cooling
Likes
0
Model age
5mo ago
created 2026-04-26
Downloads over time
Now2.6K→from1.1K↑140%
1K1.6K2.2K2.8K1.1K on Apr 292.6K on Oct 112.6K on Oct 6AprMayJunJulAugSepOct
Apr 29 → Oct 11 · 63 snapshots · spans 165 days

Benchmarks

Benchmark Score Source
Entertainment 1.9 UGI
Hazardous 2.4 UGI
Natural Intelligence 28.42 UGI
Political lean -12.5% UGI
Sensitive-Info 18.88 UGI
SocPol 1.5 UGI
UGI 44.25 UGI
Willingness (10) 9.5 UGI
W10-Adherence 9 UGI
W10-Direct 10 UGI
Writing 37.6 UGI

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.

llmfan46/Qwen3.6-35B-A3B-uncensored-heretic↓ 1,279 this lineage
ijwfly/Qwen3.6-35B-A3B-uncensored-heretic-mlx-8Bit↓ 72

Metadata

License
apache-2.0
Tags
transformers safetensors qwen3_5_moe image-text-to-text heretic uncensored decensored abliterated mlx mlx-my-repo conversational base_model:llmfan46/Qwen3.6-35B-A3B-uncensored-heretic

Related

Same base model
llmfan46/Qwen3.6-35B-A3B-uncensored-heretic-GGUF 29K mudler/Qwen3.6-35B-A3B-uncensored-heretic-APEX-GGUF 20K llmfan46/Qwen3.6-35B-A3B-uncensored-heretic-NVFP4-Experts-Only-GGUF 17K mradermacher/Qwen3.6-35B-A3B-uncensored-heretic-i1-GGUF 12K froggeric/Qwen3.6-35B-A3B-Uncensored-Heretic-MLX-4bit 9K
Total size
35.1 GB
Files
19
Quantizations
1
Registered
2026-08-22 13:56
Last updated on HF
2026-04-26 17:44
Source huggingface.co/ijwfly/Qwen3.6-35B-A3B-uncensored-heretic-mlx-8Bit ↗ Base model llmfan46/Qwen3.6-35B-A3B-uncensored-heretic →

Files by quantization

Auxiliary files 19 files 35.1 GB
model-00004-of-00008.safetensors 5.00 GB ed83627c download
model-00006-of-00008.safetensors 5.00 GB ed9dc702 download
model-00002-of-00008.safetensors 5.00 GB 7308e021 download
model-00003-of-00008.safetensors 4.99 GB 2a8d0ff0 download
model-00005-of-00008.safetensors 4.99 GB f16183b8 download
model-00001-of-00008.safetensors 4.96 GB 3daaa864 download
model-00007-of-00008.safetensors 4.69 GB f6cd6933 download
model-00008-of-00008.safetensors 515 MB 41a1d955 download
tokenizer.json 19.1 MB 639e352c download
model.safetensors.index.json 211 KB 849923d1 download
config.json 23.1 KB 64f6343b download
chat_template.jinja 7.58 KB a8755d82 download
.gitattributes 1.53 KB 52373fe2 download
README.md 1.27 KB af810377 download
tokenizer_config.json 1.24 KB 9de16b5b download
processor_config.json 1.16 KB 33818c7f download
preprocessor_config.json 390 B 2ea84a43 download
video_preprocessor_config.json 385 B 3ba673a5 download
generation_config.json 226 B 82e2ae8c download

README current version from Hugging Face

library_name: transformers
license: apache-2.0
license_link: https://huggingface.co/Qwen/Qwen3.6-35B-A3B/blob/main/LICENSE
pipeline_tag: image-text-to-text
tags:

  • heretic
  • uncensored
  • decensored
  • abliterated
  • mlx
  • mlx-my-repo
    base_model: llmfan46/Qwen3.6-35B-A3B-uncensored-heretic

ijwfly/Qwen3.6-35B-A3B-uncensored-heretic-mlx-8Bit

The Model ijwfly/Qwen3.6-35B-A3B-uncensored-heretic-mlx-8Bit was converted to MLX format from llmfan46/Qwen3.6-35B-A3B-uncensored-heretic using mlx-vlm version 0.4.4.

Use with mlx-openai-server

pip install mlx-openai-server

You have to choose trade-off between model types (model-type parameter):

  • lm: prefix caching works, but vision is not supported
  • multimodal: vision works, but prefix caching is unavailable
mlx-openai-server launch \
  --model-path ijwfly/Qwen3.6-35B-A3B-uncensored-heretic-mlx-8Bit \
  --reasoning-parser qwen3_5 \
  --model-type lm \
  --tool-call-parser qwen3_coder \
  --context-length 65535 \
  --port 8432 \
  --temperature 0.6 \
  --top-p=0.95 \
  --top-k=20 \
  --min-p=0.0 \
  --presence-penalty=0.0 \
  --repetition-penalty=1.0

README history 2 versions

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

  1. 2026-04-26Update README.md551a5571.3 KB
    Loading...
  2. 2026-04-26Upload folder using huggingface_hub97ea92d1.2 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