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

alexgusevski/Huihui-Qwen3-VL-4B-Instruct-abliterated-q8-mlx

alexgusevski Qwen 4.0B multimodal
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/alexgusevski%2FHuihui-Qwen3-VL-4B-Instruct-abliterated-q8-mlx"
Response includes
  • classification m1
  • files 15
  • benchmarks 11 entries
  • hub_downloads_all_time 186
  • author_summary 64 models
  • readme_text full
Try next
GET/models?author=alexgusevski
All 64 models by alexgusevski
 GET/models?family=qwen&method=m1
Sibling qwen models · same method
 GET/methods/m1
What "m1" means · taxonomy
10 credits · hourly refresh · ~4 KB payload Get an API key →
Abliteration classifier · v1.0.0
M1
Primary method

Direct removal

No other method signals detected in this model.
Confidence
MEDIUM
Why this label 3 signals
Method inferred from partial signals - repository name, related files, or tag patterns. Producer identity not confirmed; label may sharpen or shift as we gather more evidence.
  • 'abliterated' in name/tags
  • is_gguf=0 (base model)
  • no specific method indicators - defaulting to M1 (most common)
How we label → About M1 →
Refusal direction extracted via
Extraction technique

Difference-of-means

FromArditi et al. (2024) ↗
Confidence
MEDIUM
Why we say so
primary_method=M1; difference-of-means is the reference extraction for M1/M3 (Arditi 2024)
What is a refusal direction? → Read the paper ↗
Downloads · lifetime
186
15 last 30d - cooling
Likes
0
Model age
9mo ago
created 2026-01-10
Downloads over time
Now195→from0↑0%
0721432150 on Jan 7195 on Oct 11195 on Oct 8JanMarMayJulSep
Jan 7 → Oct 11 · 79 snapshots · spans 277 days

Benchmarks

Portrait before abliteration
Benchmarks of the base model as it stood before the refusal-removal operation. Compare with the numbers above to see what the operation cost.
Benchmark Score Source
Entertainment 1.3 UGI
Hazardous 1.8 UGI
Natural Intelligence 12.82 UGI
Political lean -16.0% UGI
Sensitive-Info 13.07 UGI
SocPol 1 UGI
UGI 34.55 UGI
Willingness (10) 7.8 UGI
W10-Adherence 8.5 UGI
W10-Direct 7 UGI
Writing 12.55 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.

Qwen/Qwen3-VL-4B-Instruct this lineage
alexgusevski/Huihui-Qwen3-VL-4B-Instruct-abliterated-q8-mlx↓ 15

Variants by this author 6 formats · 133 downloads combined

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

Metadata

License
apache-2.0
Tags
transformers safetensors qwen3_vl image-text-to-text abliterated uncensored mlx conversational base_model:Qwen/Qwen3-VL-4B-Instruct base_model:quantized:Qwen/Qwen3-VL-4B-Instruct license:apache-2.0 endpoints_compatible

Related

Other by alexgusevski
Huihui-Qwen3-VL-8B-Instruct-abliterated-q8-mlx 437 Llama-3.2-8X3B-MOE-Dark-Champion-Instruct-uncensored-abliterated-18.4B-mlx-8Bit 347 Huihui-Qwen3-VL-8B-Instruct-abliterated-q4-mlx 278 Mistral-Nemo-Inst-2407-12B-Thinking-Uncensored-HERETIC-HI-Claude-Opus-mlx-fp16 263 Llama3.3-8B-Instruct-Thinking-Heretic-Uncensored-Claude-4.5-Opus-High-Reasoning-mlx-2Bit 205
Same base model
huihui-ai/Huihui-Qwen3-VL-4B-Instruct-abliterated 15K huihui-ai/Huihui-Qwen3-VL-4B-Instruct-abliterated-FP8 459 Goekdeniz-Guelmez/Josiefied-Qwen3-VL-4B-Instruct-abliterated-beta-v1 380 Felldude/Qwen3-VL-4B-Instruct-Uncensored-FP8 293 Felldude/Qwen3-VL-4B-Instruct-Uncensored 151
Total size
4.75 GB
Files
15
Quantizations
1
Registered
2026-08-22 13:56
Last updated on HF
2026-01-10 12:01
Source huggingface.co/alexgusevski/Huihui-Qwen3-VL-4B-Instruct-abliterated-q8-mlx ↗ Base model Qwen/Qwen3-VL-4B-Instruct ↗

Files by quantization

Auxiliary files 15 files 4.77 GB
model.safetensors 4.75 GB b85b2d42 download
tokenizer.json 10.9 MB 05d47c87 download
vocab.json 2.65 MB 4783fe10 download
model.safetensors.index.json 97.6 KB f11e759c download
chat_template.jinja 5.17 KB 12438680 download
config.json 2.01 KB 2c5f4c5e download
processor_config.json 1.88 KB fdc6f993 download
.gitattributes 1.53 KB 52373fe2 download
tokenizer_config.json 1.04 KB d5f8164a download
video_preprocessor_config.json 817 B e32b1d90 download
preprocessor_config.json 782 B 2fa65535 download
README.md 755 B 550c140d download
added_tokens.json 707 B b54f9135 download
special_tokens_map.json 613 B ac23c0aa download
generation_config.json 213 B bdb4e037 download

README current version from Hugging Face

license: apache-2.0
base_model:

  • Qwen/Qwen3-VL-4B-Instruct
    pipeline_tag: image-text-to-text
    tags:
  • abliterated
  • uncensored
  • mlx
    library_name: transformers

alexgusevski/Huihui-Qwen3-VL-4B-Instruct-abliterated-q8-mlx

This model was converted to MLX format from huihui-ai/Huihui-Qwen3-VL-4B-Instruct-abliterated using mlx-vlm version 0.3.10.
Refer to the original model card for more details on the model.

Use with mlx

pip install -U mlx-vlm

python -m mlx_vlm.generate --model alexgusevski/Huihui-Qwen3-VL-4B-Instruct-abliterated-q8-mlx --max-tokens 100 --temperature 0.0 --prompt "Describe this image." --image <path_to_image>

README history 1 version

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

  1. 2026-01-10Upload folder using huggingface_hub9bfa45e755 B
    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