# Huihui-gemma-4-12B-it-abliterated

> Solid mid-size Gemma 4 abliterated variant from the most active publisher.


- **Base model:** google/gemma-4-12b-it
- **Family:** Gemma
- **Size:** 12B
- **Method:** abliteration
- **Approx. VRAM (Q4):** ~8 GB
- **License:** Apache-2.0 / Gemma
- **Quantizations:** GGUF, Transformers
- **Last updated:** 2026-11-11
- **Hugging Face:** https://huggingface.co/huihui-ai/Huihui-gemma-4-12B-it-abliterated
- **Ollama:** `ollama run huihui_ai/gemma-4-abliterated:12b`

## Notes

Solid mid-size Gemma 4 abliterated variant from the most active publisher.

Canonical page: https://www.wahloon.net/models/huihui-gemma-4-12b-abliterated

You are responsible for how you use this model.