Kie.ai Generator
AI-generatedSummary
Generate an image from a required prompt using a Kie.ai Nano Banana or GPT Image 2 model. Optionally provide reference images and image settings; either wait for the generated image as binary output or submit the task for webhook completion.
Inputs
- Model / Mode (required) — Choose Nano Banana Pro, Nano Banana 2, GPT Image 2 text-to-image, or GPT Image 2 image-to-image. Defaults to Nano Banana Pro.
- Prompt (required) — Describe the image to generate. For described reference images, their non-empty descriptions are added as reference context before this prompt.
- Reference Images — For supported modes, provide reference image URLs as a simple list or as URL-and-description entries. GPT Image 2 image-to-image requires at least one URL; the node passes the provided URLs to Kie.ai.
- Aspect Ratio — Choose auto or a listed aspect ratio. Defaults to auto.
- Resolution — Choose 1K, 2K, or 4K. Defaults to 1K.
- Wait for Completion — Defaults to true. When enabled, the node polls for completion and downloads the first returned image. When disabled, it submits the task and returns its task ID; an optional Callback URL is included in the request when supplied.
- Advanced Settings — Set GPT Image 2 quality (low, medium, or high; defaults to high), output format (JPEG, PNG, or WebP; defaults to JPEG), and, when waiting, polling interval and timeout (defaults to 3 and 120 seconds).
Output shape
One output item per input item, with incoming JSON properties preserved. When waiting, each successful item contains JSON fields taskId, imageUrl, modelMode, aspectRatio, and success, plus the downloaded image in binary property image; incoming binary properties are preserved. When not waiting, the item contains taskId, success, status set to submitted, and callbackUrl in JSON.
The node requires Kie.ai API credentials. In wait mode it downloads only the first result URL. Errors stop execution unless Continue On Fail is enabled, in which case an error and success=false are returned for that input item. Callback mode returns submission details; completion is sent to the configured callback URL.