Video GenerationActive
ByteDance / DreamActor
DreamActor
Transform images into dynamic videos using ByteDance's DreamActor model. Supports real people, animation, and pets with facial and body movement driving.
Image to Video
Model ID
DreamActor
Provider
bytedance
Updated
1774537888
wiro playground—bytedance/DreamActor
Updated 1774537888
# Overview
DreamActor is a tool developed by ByteDance for generating videos from input images and reference videos. It enables users to create dynamic video content by leveraging facial expressions and body movements from source materials.
# What you can build
- Animated character videos
- Video editing enhancements
- Motion graphics for content creators
- Personalized video content generation
# Inputs
- **Input Image**: JPEG, JPG, or PNG format (max 4.7 MB). Resolution between 480×480 and 1920×1080.
- **Video**: MP4 or WebM format (max 30 seconds). Resolution between 200×200 and 2048×1440.
# Outputs
- Generated video with facial expressions and body movements driven by input data
# Recommended settings
- Use high-resolution input images (minimum 480×480)
- Ensure reference videos contain clear facial and body movements
- Test with various subject types including real people, animals, and animations
# Limitations
- Overly abstract subjects may fail to generate
- Video duration must not exceed 30 seconds
- Resolution of input images beyond 1920×1080 will be proportionally reduced
# Safety & compliance
- This tool is designed for creative and entertainment purposes
- Users should ensure proper permissions when using real person references
- Generated content should comply with platform guidelines
API quick start
Run DreamActor with a single API call.
POST https://api.wiro.ai/v1/Run/ByteDance/DreamActor
{
"inputImage": "https://your-cdn.com/input.png",
"inputVideo": "https://your-cdn.com/input.mp4"
}