What are the hardware requirements for Gemma 3 12B Janus Abliterated V2 Base And GGUF?

Gemma 3 12B Janus Abliterated V2 Base And GGUF requires approximately 24.2 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Gemma 3 12B Janus Abliterated V2 Base And GGUF and how large is it?

Gemma 3 12B Janus Abliterated V2 Base And GGUF is developed by SOMEHOTMEAL, a model with 12b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate Gemma 3 12B Janus Abliterated V2 Base And GGUF?

Gemma 3 12B Janus Abliterated V2 Base And GGUF is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Gemma 3 12B Janus Abliterated V2 Base And GGUF by SOMEHOTMEAL — VRAM 24.2GB

Name: Gemma 3 12B Janus Abliterated V2 Base And GGUF
Author: SOMEHOTMEAL

Gemma 3 12B Janus Abliterated V2 Base And GGUF is an open-source language model by SOMEHOTMEAL. Features: 12b LLM, VRAM: 24.2GB, Quantized, Merged, LLM Explorer Score: 0.32.

Merged Model Arxiv:2311.03099 Base model:dreamfast/gemma-3-1... Base model:google/gemma-3-12b-... Conversational Endpoints compatible Fp16 Gemma3 Gguf Image-text-to-text Q2 Q8 0 Quantized Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/SOMEHOTMEAL/Gemma-3-12B-Janus-Abliterated-V2-Base_And_GGUF

Gemma 3 12B Janus Abliterated V2 Base And GGUF Parameters and Internals

LLM Name	Gemma 3 12B Janus Abliterated V2 Base And GGUF
Repository 🤗	https://huggingface.co/SOMEHOTMEAL/Gemma-3-12B-Janus-Abliterated-V2-Base_And_GGUF
Base Model(s)	Gemma 3 12B It Heretic Gemma 3 12B It DreamFast/gemma-3-12b-it-heretic google/gemma-3-12b-it
Merged Model	Yes
Model Size	12b
Required VRAM	24.2 GB
Updated	2026-09-16
Maintainer	SOMEHOTMEAL
Model Type	gemma3
Model Files	4.8 GB 7.3 GB 9.7 GB 12.5 GB 23.5 GB 4.9 GB: 1-of-5 4.9 GB: 2-of-5 4.9 GB: 3-of-5 4.9 GB: 4-of-5 4.6 GB: 5-of-5
GGUF Quantization	Yes
Quantization Type	fp16\|gguf\|q2\|q8_0\|q4_k
Model Architecture	Gemma3ForConditionalGeneration
Transformers Version	5.18.0.dev0
Tokenizer Class	GemmaTokenizer
Padding Token	<pad>

Rank the Gemma 3 12B Janus Abliterated V2 Base And GGUF capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to Gemma 3 12B Janus Abliterated V2 Base And GGUF

Best Alternatives	Context / RAM	Downloads	Likes
Gemma 3 12B It GGUF	0K / 0.9 GB	64749	193
Gemma 3 12B It Qat GGUF	0K / 0.9 GB	49888	54
Gemma 3 12B It Qat Int4 GGUF	0K / 0.9 GB	1888	10
... 3 12B It Qat Q4 0 Unquantized	0K / 24.3 GB	72553	158
...emma 3 12B It Unsloth Bnb 4bit	0K / 12.8 GB	27048	24
Gemma 3 12B It Bnb 4bit	0K / 7.8 GB	7552	45
Translategemma 12B It 4bit	0K / 6.6 GB	894	4
Translategemma 12B It 8bit	0K / 12.5 GB	568	8
Gemma 3 12B It 4bit	0K / 8.1 GB	6832	6
Gemma 3 12B It Qat Bnb 4bit	0K / 7.8 GB	2396	4

Note: green Score (e.g. "73.2") means that the model is better than SOMEHOTMEAL/Gemma-3-12B-Janus-Abliterated-V2-Base_And_GGUF.