API Endpoint URL
Base URL for all API requests to this endpoint.
https://modelslab.com/api/v7/video-fusion/motion-control
API Authentication
Authentication requires a valid API key included in the request. Generate and manage your API keys from your developer dashboard. Include the key in the key parameter for all API requests.
Integration Examples
Production-ready code samples for API integration
{"prompt": "make this image accurately animate to the video","init_image": "https://assets.modelslab.com/generations/8d4def68-7edd-449c-b5ce-2ba4a5108f43.jpg","init_video": "https://assets.modelslab.com/generations/5c570825-d8dc-421c-876d-e0ce3ee6b7bc.mov","character_orientation": "image","model_id": "kling-motion-control","mode": "std","key": "YOUR_API_KEY"}
SDKs
Official SDKs
Production-ready SDKs and client libraries for all major programming languages
API Parameters
Technical specifications for API request parameters.
| Field Name | Parameter | Technical Description |
|---|---|---|
| Prompt | prompt | Text prompt words, which can include positive and negative descriptions |
| Image | init_image | Reference Image. The characters, backgrounds, and other elements in the generated video are based on the reference image The video content must meet the following requirements: Keep the character’s proportions consistent between the image and the reference video. Avoid using a full-body reference video with a half-body image. Ensure the character’s entire body and head are clearly visible and not obstructed. Character Selection is supported for up to 1 character. For best results, choose characters with larger proportions in the frame. Supports uploading images of realistic/stylised characters (including humans, humanoid animals, certain pure animals, and characters with partial humanoid body proportions). |
| Video | init_video | The video content must meet the following requirements: Ensure the character’s entire body and head are clearly visible and not obstructed. Upload a single character motion reference. For motion reference with two or more characters, the motion of the character occupying the largest portion of the frame will be used for generation. Real human actions are recommended, while certain stylised humanoid or humanoid body proportions can be recognized. Avoid cuts and camera movements in the motion reference. Avoid overly fast motions; steady, moderate movements yield the best results. The video file supports. mp4/. mov, with a file size not exceeding 100MB, and only supports side lengths between 340px and 3850px. If the above verification fails, an error code and other information will be returned. The lower limit of video duration should not be less than 3 seconds, and the upper limit of duration is related to the character orientation reference (character_orientation): When the orientation of the character matches that of the character in the video, the video can last up to 30 seconds; When the orientation of the character is consistent with the character in the picture, the video duration can reach up to 10 seconds; |
| Character Orientation | character_orientation | Use "image" when the character in your reference image and reference video are facing/posing in the same direction. This keeps the character looking the same way as in your image. Use "video" when you want the character to follow the motion and direction shown in your reference video, even if it differs from the image orientation. This allows for longer videos (up to 30 seconds). |
| Model | model_id | Model to use for generation |
| Mode | mode | Video generation mode Enum values: std, pro std: Standard Mode, which is cost-effective pro: Professional Mode, generates videos use longer duration but higher quality video output |