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

cyberCyber99/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive-OptiQ-4bit-MLX

cyberCyber99 Qwen 35B MoE 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/cyberCyber99%2FQwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive-OptiQ-4bit-MLX"
Response includes
  • classification m-uncensored
  • files 13
  • hub_downloads_all_time 4,389
  • author_summary 1 models
  • readme_text full
Try next
GET/models?family=qwen&method=m-uncensored
Sibling qwen models · same method
 GET/methods/m-uncensored
What "m-uncensored" means · taxonomy
10 credits · hourly refresh · ~4 KB payload Get an API key →
Abliteration classifier · v1.0.0
M-U
Primary method

Uncensored (method unknown)

No other method signals detected in this model.
Confidence
LOW
Why this label 3 signals
Weak or ambiguous signals. Best guess based on catalog patterns; treat as tentative and check the evidence below.
  • 'uncensored' in name/tags but no 'abliterated' marker
  • method not identifiable from author declaration alone
  • may be DPO fine-tune, prompt engineering, or unknown technique
How we label → About M-U →
Refusal direction extraction

No specific extraction method could be identified for this model. The producer either did not document it or used a proprietary pipeline.

What is a refusal direction? →
Downloads · lifetime
4K
519 last 30d - stable
Likes
2
Model age
4mo ago
created 2026-06-09
Downloads over time
Now4.6K→from365↑1,154%
1541.8K3.4K5K365 on Jun 104.6K on Oct 11JunJulAugSepOct
Jun 10 → Oct 11 · 57 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.

ericpandev/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive-bf16↓ 3,002 this lineage
cyberCyber99/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive-OptiQ-4bit-MLX↓ 519

Metadata

License
apache-2.0
Tags
mlx safetensors qwen3_5_moe optiq mixed-precision text-generation conversational base_model:ericpandev/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive-bf16 base_model:quantized:ericpandev/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive-bf16 license:apache-2.0 4-bit region:us

Related

Same base model
dawncr0w/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive-OptiQ-6bpw-MLX 1K dawncr0w/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive-text-oQ4 499 dawncr0w/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive-OptiQ-4bit-MLX 476 dawncr0w/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive-OptiQ-5bpw-MLX 315 dawncr0w/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive-oQ8 307
Total size
18.2 GB
Files
13
Quantizations
1
Registered
2026-08-22 13:56
Last updated on HF
2026-06-09 21:53
Source huggingface.co/cyberCyber99/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive-OptiQ-4bit-MLX ↗ Base model ericpandev/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive-bf16 →

Files by quantization

Auxiliary files 13 files 18.2 GB
model-00002-of-00004.safetensors 5.00 GB 8e2787a6 download
model-00003-of-00004.safetensors 4.99 GB 79e16de3 download
model-00001-of-00004.safetensors 4.98 GB a9579d37 download
model-00004-of-00004.safetensors 3.20 GB 27b5fb21 download
tokenizer.json 19.1 MB a11a3f42 download
model.safetensors.index.json 183 KB 09bd62c2 download
config.json 133 KB cab93de2 download
optiq_metadata.json 41.9 KB e5b2b4e8 download
chat_template.jinja 7.58 KB a8755d82 download
README.md 1.46 KB c7acc027 download
tokenizer_config.json 1.13 KB 8d610519 download
.gitattributes 786 B e77110be download
generation_config.json 315 B 18c7b633 download

README current version from Hugging Face

license: apache-2.0
library_name: mlx
pipeline_tag: text-generation
tags:

  • mlx
  • optiq
  • mixed-precision
  • text-generation
    base_model: ericpandev/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive-bf16

Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive OptiQ 4bit

Source: ericpandev/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive-bf16
Original: HauhauCS/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive

Quantization

Field Value
Method OptiQ mixed-precision MLX
Target BPW 4
Achieved BPW 4.07
Candidate bits 4, 8
Group size 64
Calibration optiq (24 samples)
Model size 18.17 GiB
Layer allocation 14 high-bit / 377 low-bit
Layer policy entries 391

Notes

  • This is a text-only MLX export; vision components are omitted.
  • Native MTP is disabled for this build.
  • The original model is multimodal; this repository contains only the language model weights.
  • Built from the bf16 source snapshot listed above.

Usage

huggingface-cli download dawncr0w/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive-OptiQ-4bit-MLX --local-dir Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive-OptiQ-4bit-MLX
python -m mlx_lm.generate --model Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive-OptiQ-4bit-MLX --prompt "Write a short greeting."

README history 1 version

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

  1. 2026-06-09Duplicate from dawncr0w/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive-OptiQ-...d6de0ec1.5 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