Models

Nano Banana 2 Image Edit

Edits up to 4 source images on Gemini 3.1 Flash Image from a prompt, with 1K, 2K or 4K output in ten aspect ratios.

Cost depends on your settings.

Playground

Input
Advanced settings

Adjust the randomness of generated text.

Default: 40

Output
Nano Banana 2 Image Edit: Risograph hood with a visor readout
Example

Try an example

Call this model from your code

Sends your Playground inputs to Comfy Router.

API docs (opens in a new tab)
  1. Get an API key

    Needs a paid Cloud plan with credits.

  2. Install the SDK

    Python 3.10+. Also on npm as @comfyorg/sdk.

    pip install comfy-sdk==0.2.0
  3. Run it

    Calls Comfy Router at this endpoint with your COMFY_API_KEY.

    POST /v2/models/vertexai/gemini-3.1-flash-image

Fill in the Playground inputs to generate a request.

More from Google

Compare Google's other models. Price is the default setting; resolution and duration show the full range each model offers.