Models

Gemma 4 26B-A4B uncensored MoE: 1M context, vision, fast. ~91% recall, honestly documented.

Gemma 4 26B MoE quantized by BatiAI. 77 t/s on M4 Max. Requires 24GB+ Mac.

Gemma 4 26B MoE (Google DeepMind) with thinking mode enabled. Mixture-of-Experts — 25.2B total / 3.8B active parameters, 256K context. Supports text and image input. Knowledge cutoff: January 2025.

Gemma 4 26B MoE (Google DeepMind) with thinking mode disabled. Mixture-of-Experts — 25.2B total / 3.8B active parameters, 256K context. Supports text and image input. Knowledge cutoff: January 2025.

😈 Uncensored Gemma 4 26B (A4B MoE) for Ollama. 👁️ Native Vision, 🛠️ Tool Calling, GGUF, Rust 1.99.1 , Linux, Windows AI Coding Agents and multimodal local AI development. 🚀

Gemma 4 26B A4B IT MM QAT Unsloth Dynamic 2.0 GGUF MoE

Gemma 4 26B A4B IT QAT Unsloth Dynamic 2.0 GGUF MoE

A fast, capable and reduced-refusal Gemma 4 MoE for creative work, coding and unrestricted local exploration.

Unsloth's Gemma4 QAT models with enforced context windows.