Nano-banana via ComfyUI Partner Nodes!
Nano-banana aka Gemini-2.5-Flash Image is here!
ComfyUI now supports nano-banana via our native Partner Nodes — letting you plug cutting-edge generation directly into your workflows. Everything runs inside your ComfyUI canvas.
Model Highlights
Impressive character consistency and fidelity
Supports both text-to-image & image-to-image generation
The model can provide text output at the same time
Customizable resolutions and aspect ratios
Getting Started
Update ComfyUI to the nightly ( Stable and Desktop will be ready soon)
Double click on the canvas, search for “Goolge Gemini Image" node
Bring the workflow to your canvas and run!
Example Showcase
Chain Nano-banana with Wan 2.2
Token Cost
The model runs via API credits. Cost depends on the input and output token numbers, consistent with the model provider’s pricing.
As alway, enjoy creating!
Write back if anyone tests it. What about censorship and other stuff? Is it the same model as on LM Arena or the one in Google Ai Sudio?
wow, it's cool. i find another site can use nanobanana model:
https://nanobananas.site/