Meet Nano Banana Pro in ComfyUI!

4K generation, crazy text rendering and character consistency

Today, we’re excited to bring Nano Banana Pro (Gemini 3 Pro Image) into ComfyUI!

Nano Banana Pro is Google DeepMind’s new flagship model for high-fidelity image generation and editing. It pushes beyond casual creation into production-ready visuals.

Highlights of Nano Banana Pro

• World knowledge: Generates accurate, real-world images by tapping into Search’s knowledge base.

• Text & translation: Renders clean text and supports detection/translation across 10 languages.

• High resolutions and studio controls: Blend up to 14 images, adjust angles and focus, apply color grading, and create in native 4K.

• Consistency: Turns sketches into dresses or blueprints into 3D visuals with high-fidelity consistency.

Get Started

For new users, the easiest way to use Nano Banana 2 is to go to Comfy Cloud through the link below. No extra setup needed.

For local users, please update ComfyUI and go to Template → Nano Banana Pro!

The workflow will look like this:

Examples

Text Rendering

Understanding of materials

Multiple-Object Reference & Series Editing

Character Consistency

Pricing

• $0.134 per 1K & 2K image output

• $0,24 per 4K image output

What’s more?

You can now also chain multiple Nano Banana Pro nodes or any Partner Nodes in one workflow to achieve multi-stage creation.

We have also added support for Gemini 3, Google’s latest language model in ComfyUI. Many ComfyUI users have been using the language models to help with image captioning and prompt generation. Please search for the “Google Gemini” node in ComfyUI directly!

Please visit our Documentation for detailed pricing about text and thinking tokens.
As always, enjoy creating!