LLM EXPLORER 62,323 MODELS INDEXED

Supergemma4 26B Abliterated Multimodal by Jiunsong

By Jiunsong · 354 downloads

Supergemma4 26B Abliterated Multimodal is an open-source language model by Jiunsong. Features: 26b LLM, VRAM: 51.6GB, License: gemma, Instruction-Based, LLM Explorer Score: 0.26.

  Bf16   Conversational   En   Endpoints compatible   Gemma4   Image-text-to-text   Instruct   Instruction-following   Ko   Multimodal   Region:us   Safetensors   Sharded   Tensorflow   Tool-use

Supergemma4 26B Abliterated Multimodal Parameters and Internals

LLM NameSupergemma4 26B Abliterated Multimodal
Repository πŸ€—https://huggingface.co/Jiunsong/supergemma4-26b-abliterated-multimodal 
Base Model(s)  google/gemma-4-26b-it   google/gemma-4-26b-it
Model Size26b
Required VRAM51.6 GB
Updated2026-09-27
MaintainerJiunsong
Model Typegemma4
Instruction-BasedYes
Model Files  5.3 GB: 1-of-11   4.9 GB: 2-of-11   4.9 GB: 3-of-11   4.9 GB: 4-of-11   4.9 GB: 5-of-11   4.9 GB: 6-of-11   4.9 GB: 7-of-11   4.9 GB: 8-of-11   4.9 GB: 9-of-11   4.9 GB: 10-of-11   2.2 GB: 11-of-11
Supported Languagesen ko
Model ArchitectureGemma4ForConditionalGeneration
Licensegemma
Transformers Version5.5.0.dev0
Tokenizer ClassGemmaTokenizer
Padding Token<pad>

Best Alternatives to Supergemma4 26B Abliterated Multimodal

Best Alternatives
Context / RAM
Downloads
Likes
G4 MeroMero 26B A4B0K / 51.6 GB650100
Gemopus 4 26B A4B It0K / 51.6 GB128214
Gemopus 4 26B A4B It Preview0K / 51.6 GB12432
G4 26B A4B Musica V10K / 51.6 GB3816
Gemopus 4 26B A4B It GGUF0K / 1.2 GB2783101
Note: green Score (e.g. "73.2") means that the model is better than Jiunsong/supergemma4-26b-abliterated-multimodal.