KimiInstall

Summary

Sends a vision-enabled chat request to the Kimi (Moonshot) API, combining a text prompt with an image or video input.

Inputs

• items — Any incoming items; when using binary image source, each item must contain a binary property (default 'data') with base64-encoded data and optional mimeType.

Output shape

array

One output object per input item, each containing the model response metadata and content. Fields include: model (string), usage (object), id (string), created (number), content (string), reasoning (string or null), raw (full API response).

Examples

Example 1: Describe an image from a URL

Set operation to Vision Chat, model to moonshot-v1-vision, visionPrompt to 'What is in this image?', mediaType to image, imageSource to url, imageUrl to 'https://example.com/photo.jpg'

Links

• n8n community nodes installation guide