Firefly ServicesInstall

Summary

Generate image variations asynchronously from a text prompt using Adobe Firefly's AI models.

Inputs

• Prompt (required) — Text description (1-1024 characters) guiding the image generation.
• Model Version (required) — Select the Firefly model version for generation (e.g., Image 4 Standard).
• Additional Options — Optional parameters to customize generation, including content style (Art or Photo), image size, negative prompt, number of variations (1-4), seeds for reproducibility, structure and style reference images with adherence strength, upsampler type (for certain models), and visual intensity (contrast, shadow, hue adjustments). For custom models, Custom Model ID is required.

Output shape

A JSON object representing the asynchronous job submission result, typically containing a job ID and initial metadata.

The job runs asynchronously, so the output only confirms the job submission. Use the 'Get Job Status' operation with the returned job ID to poll for completion and retrieve generated images. The operation validates prompt length and custom model ID requirements as per inputs.

Examples

Example 1: Generate 3 photographic variations of a futuristic cityscape at sunset using 'Image 4 Standard' model with default image size 2048x2048.

Prompt: 'A futuristic cityscape at sunset'; Model Version: 'image4_standard'; Number of Variations: 3; Content Class: 'photo'.