← All Models

Gemma 4 26B

26B MoE (4B active) Min 18GB VRAM Full 52GB VRAM chatcodingmultimodalresearch

Google's Gemma 4 26B multimodal MoE model. The Ollama q4 tag is 18GB, the bf16 variant is 52GB, and it supports text + image input with a 256K context window.

Run with Ollama

ollama run gemma4:26b

Which GPUs can run this?

Check my GPU →
📚 Learn to fine-tune Gemma 4 26B — Pluralsight