Maia Router icon

Maia Router

Interact with Maia Router API for AI model routing

Actions6

Image → Edit Image

AI-generated

Summary

Edit an existing image by applying modifications specified in a prompt using an AI image editing model. Supports input images provided either as binary file data or via an image URL, and returns the edited images with associated metadata and binary data.

Inputs

  • Model (required) — Select the AI image model to use for editing: GPT Image 1, Gemini Nano Banana, or Gemini Nano Banana Pro.
  • Prompt (required) — Instruction describing the desired edits to apply to the input image.
  • Input Data Mode (required) — Choose whether to provide the input image as a binary file from previous nodes or as an image URL.
  • Binary Property (required) — If using binary file input mode, specify the binary property name containing the image data.
  • Image URL (required) — If using image URL input mode, provide the URL of the image to edit.
  • Aspect Ratio — (For Gemini models only) Define the aspect ratio for the edited image output.
  • Quality — (For GPT Image 1 model only) Specify the desired quality level: Auto, High, Medium, or Low.
  • N (Number of Images) — (For GPT Image 1 model only) Number of edited images to generate, between 1 and 10.
  • Size — (For GPT Image 1 model only) Resolution size of the output images: 256x256, 512x512, or 1024x1024.
  • Size — (For Gemini models only) Choose the output image size: 1K, 2K, or 4K.

Output shape

A list of one or more edited image items, each including metadata and optionally binary image data for direct use in workflows.

Returns a JSON object with details about the edited images, including URL and base64-encoded image data converted to binary. Binary properties are named using 'edited_image_X' where X is the 1-based index. Supports different response formats for Gemini and GPT models. If the image input is invalid or missing, the operation errors out. This operation requires Maia Router API credentials and sends requests to the API's image edits endpoint or chat completions endpoint (for Gemini models). Output includes usage metadata and operation details.

Examples

Example 1: Edit an image binary file with GPT Image 1 model

Set Model to 'GPT Image 1'. Set Input Data Mode to 'Binary File' and specify the binary property containing the image. Provide an edit prompt describing changes. Optionally, configure quality, size, and number of images (N).

Example 2: Edit an image from a URL using Gemini Nano Banana

Set Model to 'Gemini Nano Banana'. Set Input Data Mode to 'Image URL' and provide the image URL. Specify the edit prompt and optionally an aspect ratio and output size (1K, 2K, 4K).

Links

Discussion