Lip Sync

Synchronize video avatar lips with audio input for natural-looking talking head videos.

Try it out

Test the Lip Sync API directly in your browser with our interactive playground.

Overview

The Lip Sync API uses advanced AI to precisely match lip movements with speech audio, creating natural and convincing talking head videos.

API Reference

Endpoint

POST https://api.tryvinci.com/vincils

Headers

Name	Type	Description
X-User-ID	string	Required. Your user ID

Request Body

Multipart form data with:

Parameter	Type	Description
audio	File	Audio file containing speech
video	File	Video file to synchronize

Response

{
  url: string,        // URL to the synchronized video
  media_type: string, // "video/mp4"
  user_id: string    // The user ID used for the request
}

Code Examples

Python

import requests

url = "https://api.tryvinci.com/vincils"
headers = {"X-User-ID": "your-user-id"}
files = {
    "audio": open("speech.wav", "rb"),
    "video": open("video.mp4", "rb")
}

response = requests.post(url, headers=headers, files=files)
result = response.json()
print(f"Synchronized Video URL: {result['url']}")

JavaScript

const formData = new FormData();
formData.append('audio', audioFile);
formData.append('video', videoFile);

const response = await fetch('https://api.tryvinci.com/vincils', {
  method: 'POST',
  headers: {
    'X-User-ID': 'your-user-id',
  },
  body: formData,
});

const result = await response.json();
console.log(`Synchronized Video URL: ${result.url}`);

Input Requirements

Video Requirements

• Format: MP4, MOV
• Resolution: 720p or higher recommended
• Clear, front-facing view of the speaker

Audio Requirements

• Format: WAV, MP3, M4A
• Clear speech with minimal background noise
• Matching duration with video

Duration Limits

• Optimal length: 5-60 seconds per request
• Longer videos should be split into segments

Best Practices

Video Quality

Use high-quality, well-lit video footage.

Audio Clarity

Ensure clear speech audio for best sync.

Performance

For best results, ensure the speaker's face is clearly visible and well-lit throughout the video.