Generate Image
AI-generatedSummary
Generates an image via the Qwen Image model from ModelScope, polls for task completion within a specified timeout, and returns the generated image binary along with task metadata.
Inputs
- main (required) — Input items; each item generates one image using the provided prompt.
Output shape
main
One output item per input. The JSON output contains task metadata; the binary output contains the generated image as base64-encoded data.
Examples
Example 1: Generate a single image with default settings.
Use the default prompt placeholder, leave negative prompt empty, keep size at 1024x1024, and use default polling intervals.