Skip to main content

AI Generate Image

Controlled node

Overview

The AI Generate Image node creates images from text prompts using an AI image generation model. It accepts a text prompt and various generation parameters to produce a Jimp image object that can be processed by other image nodes or saved to the library.

By default, the node generates a 1024x1024 image using a seed of 1 for reproducibility. The output is a Jimp image object compatible with the platform's image processing nodes.

Inputs

InputTypeDescriptionDefault
RunEventTriggers the image generation.-
PromptTextThe text description of the image to generate.-
WidthNumberThe width of the generated image in pixels.1024
HeightNumberThe height of the generated image in pixels.1024
GuidanceNumberControls how closely the image follows the prompt (0-10). Higher values enforce stricter adherence to the prompt.7
SeedNumberSets the random seed for reproducible generation.1
StepsNumberThe number of inference steps to run (minimum 1).1

Outputs

OutputTypeDescription
OutputDataThe generated image as a Jimp image object.
DoneEventFires when the image generation is complete.

Runtime Behavior and Defaults

When the Run event is triggered, the node validates and clamps all numeric parameters to valid ranges:

  • Width and Height are clamped between 32 and 1024 pixels. If invalid, they default to 256.
  • Guidance is clamped between 0 and 10. If invalid, it defaults to 1.
  • Seed must be at least 1. If invalid, it defaults to 1.
  • Steps must be at least 1. If invalid, it defaults to 1.

The node uses the Google Flash Image 2.5 model for generation. The resulting image is output as a Jimp image object, which can be directly connected to image processing nodes (such as Resize Image, Crop Image, or Composite Image) or saved to the library using the Write To Library node with file type jpeg.

Example Usage

To generate a custom image and save it to the library:

  1. Connect a Start node or any event trigger to the Run input.
  2. Provide a text prompt describing the desired image (e.g., "A futuristic cityscape at sunset").
  3. Optionally adjust the Width, Height, Guidance, Seed, or Steps parameters.
  4. Connect the Output to a Write To Library node's Data input.
  5. Set the Write To Library node's File Type property to jpeg and provide a Name and Path.
  6. Connect the Done event to trigger the write operation or subsequent workflow steps.