Models

gemma4

Gemma 4 models are designed to deliver frontier-level performance at each size. They are well-suited for reasoning, agentic workflows, coding, and multimodal understanding.

embeddinggemma-2

EmbeddingGemma 2 is a multimodal embedding model from Google built on the Gemma 4 architecture.

medgemma1.5

MedGemma 1.5 4B is an updated version of the MedGemma 4B model.

mgraffam/gemma4-agentic-fable-abliterated

Gemma4 26b-A4b uncensored, with Fable 5 distillation aimed at agentic work.

mgraffam/gemma4-opus4.8-heretic

Gemma4 26b-24b abliterated fine-tuned on Opus 4.8 reasoning

mgraffam/goetia-absolute-heretic

Merge of several abliterated 26b-A4B Gemma4 models.

ssfdre38/gemma4-turbo

Gemma4-turbo is part of the G4Turbo.com family to try and bring the Gemma 4 Model to everyone. Please visit https://g4turbo.com/ for more information about what I am doing.

iaprofesseur/SuperGemma4-26b-uncensored-Q4

Model by Jiun Song on huggingFace : Jiunsong/supergemma4-26b-uncensored-gguf-v2 (Q4)

ssfdre38/gemma4-nano

Gemma4-nano is part of the G4Turbo.com family to try and bring the Gemma 4 Model to everyone. Please visit https://g4turbo.com/ for more information about what I am doing.

cmsmanhattan/JiRackTernaryGemma4-26b-q3

cryptidbleh/gemma4-claude-sonnet-4.6

Gemma4 e2b distilled with Claude Sonnet 4.6

ttempvnn/HauhauCS-Gemma4-26B-A4B-Uncensored-HauhauCS-Balanced-Q4-K-M

SetneufPT/Gemma4-12B-IT-QAT_Q4_64K_16GB-GPU

Custom model for coding with agents to use locally with 16gb or 2x8gb GPUs (working fine...)

cmsmanhattan/JiRackTernaryGemma4-26b-q4

ttempvnn/Gemma4-26B-A4B-Uncensored-HauhauCS-Balanced-Q4-K-M

ducquoc/gemma4-fast-sonnet

Gemma4 e2b distilled with Claude Sonnet 4.6 data, fast response (low/med thinking).

justingtzk/gemma-4-26B-A4B-it-qat-GGUF

Unsloth's Gemma4 QAT models with enforced context windows.

maxwellb/gemma4-12b-it-dn

Quantized GGUF of hf.co/DuoNeural/Gemma4-12B-IT-Abliterated

justingtzk/gemma-4-31B-it-qat-GGUF

Unsloth's Gemma4 QAT models with enforced context windows.

efso/gemma-4-e4b-it-gr-v2

Gemma4GR v3: Gemma 4 E4B fine-tuned for cleaner Greek (stray foreign characters in ~1% of answers vs ~40% for stock). v2 tags were the base model (export bug), now fixed. Not a source of facts.