Firefly ServicesInstall

Summary

Generate composite images by combining a source image with AI-generated content based on a text prompt, producing one or multiple variations asynchronously.

Inputs

• Source Image Upload ID (required) — The upload ID from the storage API identifying the source image to use for composition.
• Prompt (required) — Text description (1-1024 characters) that guides the AI in creating the composite scene around the source image.
• Composite Options — Optional settings controlling the composite generation, including style (Photo or Art), output size, number of variations (1-4), alignment and insets of the source image, optional mask image upload ID, seeds for variation consistency, style reference image, and style adherence strength.

Output shape

An array of JSON objects representing the generated composite images, each typically including metadata such as generated image URLs or upload IDs corresponding to each variation.

The number of generated image variations is controlled via 'Number of Variations' (default 1, max 4). Seeds can be provided to reproduce consistent results and must match the number of variations if specified. Optional mask and style reference parameters allow refined control over the output style and composition.