Video Generation

Generate AI-powered videos using models like Seedance through TopRouter's API. Support for text-to-video and image-to-video generation.

Overview

TopRouter provides access to state-of-the-art video generation models. Create short video clips from text prompts or reference images.

Seedance Video Generation

Text-to-Video

Generate a video from a text description:

curl https://toprouter.cc/videos/generations \
  -H "Authorization: Bearer sk-your-toprouter-key" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "seedance/seedance-1.0",
    "prompt": "A golden retriever running through a meadow of wildflowers in slow motion, cinematic lighting",
    "duration": 5
  }'

Image-to-Video

Animate a reference image:

curl https://toprouter.cc/videos/generations \
  -H "Authorization: Bearer sk-your-toprouter-key" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "seedance/seedance-1.0",
    "prompt": "Gentle camera zoom with particles floating",
    "image_url": "https://example.com/reference-image.jpg",
    "duration": 5
  }'

Parameters

Parameter	Type	Required	Description
model	string	✅	Video model ID (e.g., seedance/seedance-1.0)
prompt	string	✅	Text description for the video
image_url	string	❌	Reference image URL (for image-to-video)
duration	integer	❌	Video duration in seconds
resolution	string	❌	Output resolution

Response

Video generation is typically asynchronous. The API returns a task ID that you can poll for completion:

{
  "id": "video-task-abc123",
  "status": "processing",
  "created": 1717500000
}

Once complete:

{
  "id": "video-task-abc123",
  "status": "completed",
  "video_url": "https://example.com/generated-video.mp4",
  "duration": 5
}

Tips for Better Results

• Be specific — Describe the scene, camera movement, lighting, and style
• Use cinematic language — Terms like "slow motion", "dolly shot", "aerial view" improve results
• Keep prompts concise — Focus on key visual elements
• Use reference images — Image-to-video produces more consistent results

WARNING

Video generation may take 1-5 minutes depending on the model and settings. Use asynchronous polling to check the status.

Available Models

Model	Features
seedance/seedance-1.0	Text-to-video, Image-to-video, High quality

INFO

Video generation models and capabilities are actively expanding. Check the API for the latest available models.