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

mlx-community/Qwen2.5-VL-3B-Instruct-abliterated

mlx-community Qwen 3.8B 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/mlx-community%2FQwen2.5-VL-3B-Instruct-abliterated"
Response includes
  • classification m1
  • files 15
  • benchmarks 11 entries
  • hub_downloads_all_time 842
  • author_summary 207 models
  • readme_text full
Try next
GET/models?author=mlx-community
All 207 models by mlx-community
 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
842
82 last 30d - cooling
Likes
1
Model age
20mo ago
created 2025-02-16
Downloads over time
Now882→from4↑21,950%
03236479704 on Feb 12, 2025882 on Oct 11Feb '25May '25Aug '25Nov '25FebMayAug
Feb 12, 2025 → Oct 11 · 126 snapshots · spans 606 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 0.6 UGI
Hazardous 1.8 UGI
Natural Intelligence 9.33 UGI
Political lean -15.5% UGI
Sensitive-Info 8.85 UGI
SocPol 0.5 UGI
UGI 20.07 UGI
Willingness (10) 4.2 UGI
W10-Adherence 3.5 UGI
W10-Direct 5 UGI
Writing 20.43 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/Qwen2.5-VL-3B-Instruct this lineage
mlx-community/Qwen2.5-VL-3B-Instruct-abliterated↓ 82

Metadata

Languages
en
Tags
transformers safetensors qwen2_5_vl image-text-to-text multimodal abliterated uncensored mlx conversational en base_model:Qwen/Qwen2.5-VL-3B-Instruct base_model:finetune:Qwen/Qwen2.5-VL-3B-Instruct

Related

Other by mlx-community
Qwen3.8-27B-Uncensored-OptiQ-4bit 16K Qwen3.6-40B-Claude-4.6-Opus-Deckard-Heretic-Uncensored-Thinking-8bit 7K Qwen3.8-Flash-Next-Uncensored-oQ5e-mtp 5K Josiefied-Qwen3-4B-abliterated-v1-4bit 4K Huihui-Qwen3.6-35B-A3B-Claude-4.7-Opus-abliterated-mlx-8bit 4K
Same base model
huihui-ai/Qwen2.5-VL-3B-Instruct-abliterated 22K mmp2055/Qwen2.5-VL-3B-uncensored-bughunter-v9 2K JuggaHO/Qwen2.5-VL-3B-Instruct-abliterated 660 mlx-community/Qwen2.5-VL-3B-Instruct-abliterated-4bit 105 mlx-community/Qwen2.5-VL-3B-Instruct-abliterated-8bit 62
Total size
6.99 GB
Files
15
Quantizations
1
Registered
2026-08-22 13:56
Last updated on HF
2025-02-25 21:53
Source huggingface.co/mlx-community/Qwen2.5-VL-3B-Instruct-abliterated ↗ Base model Qwen/Qwen2.5-VL-3B-Instruct ↗

Files by quantization

Auxiliary files 15 files 7.01 GB
model-00001-of-00002.safetensors 4.99 GB 88980313 download
model-00002-of-00002.safetensors 2.00 GB 0dbde998 download
tokenizer.json 10.9 MB 9c5ae00e download
vocab.json 2.65 MB 4783fe10 download
merges.txt 1.59 MB 31349551 download
model.safetensors.index.json 64.7 KB d9c73564 download
tokenizer_config.json 7.33 KB 58d34a50 download
.gitattributes 1.53 KB 52373fe2 download
config.json 1.40 KB cfb2d061 download
chat_template.json 1.03 KB 2ec1e359 download
README.md 837 B 289cf50c download
special_tokens_map.json 644 B 3a784031 download
added_tokens.json 629 B 06135f3c download
preprocessor_config.json 570 B 2bff1086 download
generation_config.json 258 B 93bd2cd2 download

README current version from Hugging Face

license_name: qwen-research
license_link: https://huggingface.co/Qwen/Qwen2.5-VL-3B-Instruct/blob/main/LICENSE
language:

  • en
    pipeline_tag: image-text-to-text
    tags:
  • multimodal
  • abliterated
  • uncensored
  • mlx
    library_name: transformers
    base_model:
  • Qwen/Qwen2.5-VL-3B-Instruct

mlx-community/Qwen2.5-VL-3B-Instruct-abliterated

This model was converted to MLX format from huihui-ai/Qwen2.5-VL-3B-Instruct-abliterated using mlx-vlm version 0.1.13.
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 mlx-community/Qwen2.5-VL-3B-Instruct-abliterated --max-tokens 100 --temp 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. 2025-02-16Upload folder using huggingface_hubf94e4e0837 B
    Loading...

Discussions 1 thread

  1. 2025-04-29PRUpdate preprocessor_config.jsonopen1 💬#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