Die Oberfläche ist lokalisiert; Skill-Beschreibungen und -Dokumente werden per KI automatisch in Ihre Sprache übersetzt. Code, Befehle, Pfade und Fachbegriffe bleiben im Original, maßgeblich ist das Dokument des Autors.
Skills entdecken
Produktivität · SKILL.md

kling-v3-pro-motion-control

kling-v3-pro-motion-control

Kling | v3 | Pro | Motion Control. Pro-quality video generation with motion from reference video and character from reference image. Triggers: motion control, kling, v3, pro, video, animation, character

Independent GitHub snapshot · Automated classification · Original documentation governs

Skill-Dokumentation

GitHub

name: kling-v3-pro-motion-control description: "Kling | v3 | Pro | Motion Control. Pro-quality video generation with motion from reference video and character from reference image. Triggers: motion control, kling, v3, pro, video, animation, character" allowed-tools: Bash(curl *), WebFetch

Kling | v3 | Pro | Motion Control

Generate high-quality videos by combining a reference image (for appearance) with a reference video (for motion). Pro tier delivers superior quality for character motion transfer.

Quick Start

Requires an each::labs API key. Get one at eachlabs.ai.

curl -X POST https://api.eachlabs.ai/v1/prediction \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $EACHLABS_API_KEY" \
  -d '{
    "model": "kling-v3-pro-motion-control",
    "version": "0.0.1",
    "input": {
      "prompt": "A character performing a ballet routine",
      "image_url": "https://example.com/ballerina.jpg",
      "video_url": "https://example.com/ballet-reference.mp4",
      "character_orientation": "video",
      "keep_original_sound": true
    }
  }'

Parameters

ParameterTypeDefaultDescription
character_orientationstringvideoControls whether output character orientation matches the reference image or video. enum: image, video
elementsarrayfalseOptional element for facial consistency binding.
image_urlstringfalseReference image URL. Characters and elements in the generated video match this image.
keep_original_soundbooleantrueWhether to keep the original sound from the reference video.
promptstringfalseText prompt describing the desired output.
video_urlstringfalseReference video URL. Character actions in the generated video follow this video.

Examples

High-quality dance motion transfer:

curl -X POST https://api.eachlabs.ai/v1/prediction \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $EACHLABS_API_KEY" \
  -d '{
    "model": "kling-v3-pro-motion-control",
    "version": "0.0.1",
    "input": {
      "prompt": "An anime character performing a dramatic sword fight sequence",
      "image_url": "https://example.com/anime-warrior.png",
      "video_url": "https://example.com/sword-fight.mp4",
      "character_orientation": "video"
    }
  }'

Character animation with image orientation:

curl -X POST https://api.eachlabs.ai/v1/prediction \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $EACHLABS_API_KEY" \
  -d '{
    "model": "kling-v3-pro-motion-control",
    "version": "0.0.1",
    "input": {
      "prompt": "A person giving a presentation and gesturing with hands",
      "image_url": "https://example.com/business-person.jpg",
      "video_url": "https://example.com/presentation-gestures.mp4",
      "character_orientation": "image",
      "keep_original_sound": false
    }
  }'

Related Models

Documentation

Gespeicherte Skills