Azure OpenAI Image icon

Azure OpenAI Image

Generate or edit images using Azure OpenAI (DALL-E 3 / gpt-image-2)

Actions2

Edit

AI-generated

Summary

Edits an input image using a text prompt, optionally masking specific areas or adding reference images.

Inputs

  • endpointType (required) — The type of Azure OpenAI endpoint you are connecting to (Classic vs. Gateway).
  • resourceName (required) — The name of your Azure OpenAI resource.
  • deploymentName (required) — The name of the model deployment (Classic endpoint only).
  • apiKey (required) — Your Azure OpenAI API Key.
  • apiVersion (required) — Azure OpenAI API version (Classic endpoint only).
  • model — The image generation model to use (Gateway endpoint only).
  • customModel — The exact model ID deployed in Azure AI Foundry (used when 'Model' is set to 'Custom').
  • prompt (required) — Text description of the image edit.
  • inputImage (required) — Name of the binary field containing the source image to edit.
  • maskImage — Optional: Name of the binary field containing a mask image (PNG/JPEG/WEBP) to guide the edit.
  • additionalImages — Optional: Comma-separated list of binary field names containing additional reference images.
  • size — The dimensions of the generated image (e.g., 1024x1024).
  • quality — The rendering quality of the image (Auto, Low, Medium, High, Standard, HD).
  • outputFormat — Format of the generated image (PNG or JPEG).
  • outputCompression — Compression level (0-100).
  • n — How many images to generate.
  • outputBinaryField — Name of the binary field to store the generated image in.

Output shape

One or more items, each containing a binary image file and metadata (prompt, size, quality, index).

Returns multiple items when 'n' > 1. Continues on failure by default, returning an error object in the JSON.

Examples

Example 1: Inpainting

Select 'Edit' operation, provide a source image in the 'Input Image Field', specify a mask image in 'Mask Image Field', and enter a text prompt describing the desired change.

Example 2: Guided Image Generation

Select 'Edit' operation, provide a source image and reference images (comma-separated) in 'Additional Image Fields', and enter a text prompt.

Links

Discussion