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-v1-pro-image-to-video

kling-v1-pro-image-to-video

Kling v1 | Pro | Image to Video. Convert images to high-quality videos using Kling v1 Pro with masking and end-frame support. Professional Kling image animation. Triggers: kling v1 pro image to video, kling pro i2v, kling v1 pro animate, kling pro image animation

Independent GitHub snapshot · Automated classification · Original documentation governs

Skill-Dokumentation

GitHub

name: kling-v1-pro-image-to-video description: "Kling v1 | Pro | Image to Video. Convert images to high-quality videos using Kling v1 Pro with masking and end-frame support. Professional Kling image animation. Triggers: kling v1 pro image to video, kling pro i2v, kling v1 pro animate, kling pro image animation" allowed-tools: Bash(curl *), WebFetch

Kling v1 | Pro | Image to Video

Convert images to high-quality videos using Kling v1 Pro. Professional-grade image-to-video with static masking for selective animation and end-frame targeting for controlled transitions.

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-v1-pro-image-to-video",
    "version": "0.0.1",
    "input": {
      "image_url": "https://example.com/cityscape.jpg",
      "prompt": "Cars driving through the streets, lights flickering, people walking on sidewalks, vibrant city life",
      "aspect_ratio": "16:9",
      "duration": 5
    }
  }'

Parameters

ParameterTypeDefaultDescription
aspect_ratiostring16:9Aspect ratio. Options: 16:9, 9:16, 1:1
cfg_scalenumber0.5Classifier-free guidance scale
durationinteger5Video duration in seconds
image_urlstring-URL of the input image
negative_promptstringblur, distort, and low qualityThings to avoid
promptstring-Text prompt for the animation
static_mask_urlstring-URL of a mask image defining static (non-animated) areas
tail_image_urlstring-URL of the image to use for the end of the video

Examples

Professional portrait animation:

curl -X POST https://api.eachlabs.ai/v1/prediction \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $EACHLABS_API_KEY" \
  -d '{
    "model": "kling-v1-pro-image-to-video",
    "version": "0.0.1",
    "input": {
      "image_url": "https://example.com/business-portrait.jpg",
      "prompt": "The person nods slightly and smiles confidently, professional office setting",
      "aspect_ratio": "9:16",
      "duration": 5,
      "negative_prompt": "blur, distort, and low quality"
    }
  }'

Selective animation with masking:

curl -X POST https://api.eachlabs.ai/v1/prediction \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $EACHLABS_API_KEY" \
  -d '{
    "model": "kling-v1-pro-image-to-video",
    "version": "0.0.1",
    "input": {
      "image_url": "https://example.com/still-life.jpg",
      "static_mask_url": "https://example.com/table-mask.png",
      "prompt": "The candle flame flickers, smoke rises gently, the table and objects remain still",
      "aspect_ratio": "16:9",
      "cfg_scale": 0.5
    }
  }'

Transition between start and end frames:

curl -X POST https://api.eachlabs.ai/v1/prediction \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $EACHLABS_API_KEY" \
  -d '{
    "model": "kling-v1-pro-image-to-video",
    "version": "0.0.1",
    "input": {
      "image_url": "https://example.com/empty-canvas.jpg",
      "tail_image_url": "https://example.com/finished-painting.jpg",
      "prompt": "An invisible hand painting a landscape, brush strokes appearing on the canvas",
      "duration": 5
    }
  }'

Related Models

Documentation

Gespeicherte Skills