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

ailexleon/gemma-4-26B-A4B-it-uncensored-heretic-mlx-vlm-8Bit

ailexleon Gemma 25B 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/ailexleon%2Fgemma-4-26B-A4B-it-uncensored-heretic-mlx-vlm-8Bit"
Response includes
  • classification m3
  • files 15
  • benchmarks 11 entries
  • hub_downloads_all_time 853
  • author_summary 15 models
  • readme_text full
Try next
GET/models?author=ailexleon
All 15 models by ailexleon
 GET/models?family=gemma&method=m3
Sibling gemma 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
853
264 last 30d - stable
Likes
0
Model age
7w ago
created 2026-08-17
Downloads over time
Now947→from255↑271%
2204867511K255 on Aug 19947 on Oct 11AugSepOct
Aug 19 → Oct 11 · 48 snapshots · spans 53 days

Benchmarks

Benchmark Score Source
Entertainment 2.8 UGI
Hazardous 3.5 UGI
Natural Intelligence 34.12 UGI
Political lean -22.2% UGI
Sensitive-Info 27.29 UGI
SocPol 2 UGI
UGI 49.03 UGI
Willingness (10) 9.2 UGI
W10-Adherence 8.5 UGI
W10-Direct 10 UGI
Writing 42.09 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/gemma-4-26B-A4B-it-uncensored-heretic↓ 389 this lineage
ailexleon/gemma-4-26B-A4B-it-uncensored-heretic-mlx-vlm-8Bit↓ 264

Metadata

License
apache-2.0
Tags
mlx safetensors gemma4 heretic uncensored decensored abliterated ara image-text-to-text conversational en base_model:llmfan46/gemma-4-26B-A4B-it-uncensored-heretic

Related

Other by ailexleon
gemma-4-12B-it-qat-uncensored-heretic-mlx-4Bit 2K gemma-4-26B-A4B-it-qat-uncensored-heretic-mlx-vlm-4Bit 1K gemma-4-31B-it-qat-uncensored-heretic-mlx-lm-4Bit 818 Huihui-Qwen3.8-27B-abliterated-mlx-4Bit 726 Huihui-Qwen3.8-27B-abliterated-mlx-8Bit 638
Same base model
llmfan46/gemma-4-26B-A4B-it-uncensored-heretic-GGUF 26K noctrex/gemma-4-26B-A4B-it-uncensored-heretic-MXFP4_MOE-GGUF 3K mradermacher/gemma-4-26B-A4B-it-uncensored-heretic-i1-GGUF 3K mradermacher/gemma-4-26B-A4B-it-uncensored-heretic-GGUF 2K ailexleon/gemma-4-26B-A4B-it-uncensored-heretic-mlx-vlm-6Bit 414
Total size
26.0 GB
Files
15
Quantizations
1
Registered
2026-08-22 13:56
Last updated on HF
2026-08-17 12:24
Source huggingface.co/ailexleon/gemma-4-26B-A4B-it-uncensored-heretic-mlx-vlm-8Bit ↗ Base model llmfan46/gemma-4-26B-A4B-it-uncensored-heretic →

Files by quantization

Auxiliary files 15 files 26.1 GB
model-00004-of-00006.safetensors 4.85 GB 2feab794 download
model-00003-of-00006.safetensors 4.85 GB 219369ed download
model-00005-of-00006.safetensors 4.85 GB 84f65cf7 download
model-00002-of-00006.safetensors 4.85 GB e9c64263 download
model-00001-of-00006.safetensors 4.83 GB 360f446c download
model-00006-of-00006.safetensors 1.82 GB 08709a25 download
tokenizer.json 30.7 MB a2619fe1 download
model.safetensors.index.json 173 KB 4c5998c9 download
chat_template.jinja 18.2 KB 4741bf6e download
config.json 11.9 KB b2c68fd7 download
tokenizer_config.json 2.75 KB dc684e39 download
.gitattributes 1.53 KB 52373fe2 download
processor_config.json 1.31 KB bb6dee0c download
README.md 795 B c9bd0a5f download
generation_config.json 217 B ed42ae71 download

README current version from Hugging Face

base_model: llmfan46/gemma-4-26B-A4B-it-uncensored-heretic
license: apache-2.0
license_link: https://ai.google.dev/gemma/docs/gemma_4_license
language: en
tags:

  • mlx
  • heretic
  • uncensored
  • decensored
  • abliterated
  • ara
    library_name: mlx
    pipeline_tag: image-text-to-text

ailexleon/gemma-4-26B-A4B-it-uncensored-heretic-mlx-vlm-8Bit

Converted to MLX format from llmfan46/gemma-4-26B-A4B-it-uncensored-heretic using mlx-vlm version 0.6.13.

Use with mlx

pip install -U mlx-vlm

python -m mlx_vlm.generate --model ailexleon/gemma-4-26B-A4B-it-uncensored-heretic-mlx-vlm-8Bit --max-tokens 100 --temperature 0.0 --prompt "Describe this image." --image <path_to_image>

README history 2 versions

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

  1. 2026-08-17Update README.md23647bd795 B
    Loading...
  2. 2026-08-17Upload folder using huggingface_hub34c97cd401 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