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

FiditeNemini/Qwen3-VL-235B-A22B-Instruct-Abliterated-3bit

FiditeNemini Qwen 235B MoE 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/FiditeNemini%2FQwen3-VL-235B-A22B-Instruct-Abliterated-3bit"
Response includes
  • classification m1
  • files 35
  • benchmarks 11 entries
  • hub_downloads_all_time 1,547
  • author_summary 4 models
  • readme_text full
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)
Refusal direction extracted via
Extraction technique

Difference-of-means

Confidence
MEDIUM
Why we say so
primary_method=M1; difference-of-means is the reference extraction for M1/M3 (Arditi 2024)
Downloads · lifetime
2K
98 last 30d - cooling
Likes
0
Model age
10mo ago
created 2025-11-27
Downloads over time
Now1.6K→from60↑2,555%
05821.2K1.7K60 on Nov 26, 20251.6K on Oct 11Nov '25JanMarMayJulSep
Nov 26, 2025 → Oct 11 · 85 snapshots · spans 319 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 2.2 UGI
Hazardous 5.9 UGI
Natural Intelligence 28.43 UGI
Political lean -22.7% UGI
Sensitive-Info 39.59 UGI
SocPol 4.6 UGI
UGI 45.56 UGI
Willingness (10) 5.8 UGI
W10-Adherence 6.5 UGI
W10-Direct 5 UGI
Writing 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.

Metadata

License
apache-2.0
Languages
en zh
Tags
mlx safetensors qwen3_vl_moe image-text-to-text conversational en zh base_model:Qwen/Qwen3-VL-235B-A22B-Instruct base_model:finetune:Qwen/Qwen3-VL-235B-A22B-Instruct license:apache-2.0 region:us

Related

Total size
111 GB
Files
35
Quantizations
1
Registered
2026-08-22 13:56
Last updated on HF
2025-11-28 03:07

Files by quantization

Auxiliary files 35 files 111 GB
model-00001-of-00022.safetensors 5.33 GB ae40feda download
model-00022-of-00022.safetensors 5.10 GB 956c5574 download
model-00019-of-00022.safetensors 5.05 GB 8f27af2b download
model-00016-of-00022.safetensors 5.05 GB 8fb103b6 download
model-00010-of-00022.safetensors 5.05 GB e1283bd6 download
model-00007-of-00022.safetensors 5.05 GB ab8b9e93 download
model-00013-of-00022.safetensors 5.05 GB ef364f66 download
model-00004-of-00022.safetensors 5.05 GB 6d14e1dd download
model-00006-of-00022.safetensors 5.01 GB 61f5ed66 download
model-00015-of-00022.safetensors 5.01 GB 6b486dda download
model-00012-of-00022.safetensors 5.01 GB 2f3f1c58 download
model-00009-of-00022.safetensors 5.01 GB 1c28d40f download
model-00021-of-00022.safetensors 5.01 GB c6ee07c0 download
model-00018-of-00022.safetensors 5.01 GB 97546a90 download
model-00003-of-00022.safetensors 5.01 GB aaab0b93 download
model-00014-of-00022.safetensors 5.01 GB c5664670 download
model-00005-of-00022.safetensors 5.01 GB c3c23fb7 download
model-00008-of-00022.safetensors 5.01 GB ff315e33 download
model-00020-of-00022.safetensors 5.01 GB 6aae3e7c download
model-00017-of-00022.safetensors 5.01 GB 97fa6825 download
model-00011-of-00022.safetensors 5.01 GB bd8e05b7 download
model-00002-of-00022.safetensors 5.01 GB 6445cbe8 download
tokenizer.json 9.71 MB 56e3bbc3 download
vocab.json 2.94 MB b6e97c78 download
merges.txt 1.59 MB e0489e76 download
model.safetensors.index.json 283 KB 2e3c165d download
tokenizer_config.json 6.25 KB 225f6209 download
chat_template.jinja 5.17 KB 12438680 download
config.json 1.99 KB e4942aa4 download
README.md 1.96 KB ddd5a5a0 download
.gitattributes 1.48 KB a6344aac download
preprocessor_config.json 462 B db77c36a download
special_tokens_map.json 362 B 6cc79df3 download
generation_config.json 180 B e3805efc download
added_tokens.json 2.00 B 9e26dfee download

README current version from Hugging Face


license: apache-2.0
language:

  • en
  • zh
    base_model:
  • Qwen/Qwen3-VL-235B-A22B-Instruct
    pipeline_tag: image-text-to-text
    library_name: mlx

QWEN3_VL_MOE (MLX Converted)

Conversion Details

  • Source: /Qwen3-VL-235B-A22B-Instruct-Abliterated/
  • Destination: /Qwen3-VL-235B-A22B-Instruct-Abliterated-3bit
  • Quantization: None (3-bit, group size 64)

Model Configuration

  • Architecture: Qwen3VLMoeForConditionalGeneration
  • Model Type: qwen3_vl_moe
  • Vocab Size: 151936
  • Context Length: 262144

Files

  • added_tokens.json (0.00 MB)
  • chat_template.jinja (0.01 MB)
  • config.json (0.00 MB)
  • generation_config.json (0.00 MB)
  • merges.txt (1.59 MB)
  • model-00001-of-00022.safetensors (5462.83 MB)
  • model-00002-of-00022.safetensors (5132.08 MB)
  • model-00003-of-00022.safetensors (5133.09 MB)
  • model-00004-of-00022.safetensors (5166.09 MB)
  • model-00005-of-00022.safetensors (5132.08 MB)
  • model-00006-of-00022.safetensors (5133.09 MB)
  • model-00007-of-00022.safetensors (5166.09 MB)
  • model-00008-of-00022.safetensors (5132.08 MB)
  • model-00009-of-00022.safetensors (5133.09 MB)
  • model-00010-of-00022.safetensors (5166.09 MB)
  • model-00011-of-00022.safetensors (5132.08 MB)
  • model-00012-of-00022.safetensors (5133.09 MB)
  • model-00013-of-00022.safetensors (5166.09 MB)
  • model-00014-of-00022.safetensors (5132.08 MB)
  • model-00015-of-00022.safetensors (5133.09 MB)
  • model-00016-of-00022.safetensors (5166.09 MB)
  • model-00017-of-00022.safetensors (5132.08 MB)
  • model-00018-of-00022.safetensors (5133.09 MB)
  • model-00019-of-00022.safetensors (5166.09 MB)
  • model-00020-of-00022.safetensors (5132.08 MB)
  • model-00021-of-00022.safetensors (5133.09 MB)
  • model-00022-of-00022.safetensors (5222.21 MB)
  • model.safetensors.index.json (0.28 MB)
  • preprocessor_config.json (0.00 MB)
  • special_tokens_map.json (0.00 MB)
  • tokenizer.json (9.71 MB)
  • tokenizer_config.json (0.01 MB)
  • vocab.json (2.94 MB)

README history 3 versions

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

  1. 2025-11-28Update README.md1f7a7c92 KB
    Loading...
  2. 2025-11-28Update README.md0735e3d1.9 KB
    Loading...
  3. 2025-11-28Upload folder using huggingface_hubdc1833e1.8 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.

Open in Abliteration