API Feature

Clone Any Voice with AI

Transform audio using AI voice cloning. Keep the emotion, change the voice.

Vydra's Voice Cloning API uses ElevenLabs' industry-leading speech-to-speech technology to transform any audio into a different voice while preserving the original emotion, pacing, and delivery.

Upload your source audio and specify the target voice, and our API handles the complex audio processing to produce natural-sounding results. Perfect for dubbing, content creation, and accessibility applications.

Why Use Our Voice Cloning API

Emotion Preservation

Our AI maintains the emotional delivery of the original audio while changing the voice characteristics.

Multiple Voices

Choose from dozens of pre-built voices or use ElevenLabs voice IDs for custom clones.

High Fidelity

44.1kHz audio output with natural prosody and minimal artifacts.

Fast Turnaround

Process audio in seconds, not minutes. Real-time streaming available for shorter clips.

Use Cases

See how developers and businesses use our voice cloning api to build amazing products.

Video Dubbing

Dub videos into different voices or languages while maintaining emotional authenticity.

Podcast Production

Create consistent voice branding across multiple hosts or transform guest recordings.

Audiobook Production

Generate different character voices from a single narrator recording.

Accessibility

Convert content into preferred voices for users with hearing or processing differences.

Simple API Integration

Get started with just a few lines of code.

cURLPOST

curl -X POST https://vydra.ai/api/v1/models/elevenlabs/sts \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "audio_url": "https://example.com/source-audio.mp3",
    "voice": "adam",
    "model_id": "eleven_multilingual_sts_v2"
  }'

JavaScript

const response = await fetch('https://vydra.ai/api/v1/models/elevenlabs/sts', {
  method: 'POST',
  headers: {
    'Authorization': 'Bearer YOUR_API_KEY',
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    audio_url: 'https://example.com/source-audio.mp3',
    voice: 'adam',
    model_id: 'eleven_multilingual_sts_v2',
  }),
});

const { audioUrl } = await response.json();

Python

import requests

response = requests.post(
    'https://vydra.ai/api/v1/models/elevenlabs/sts',
    headers={'Authorization': 'Bearer YOUR_API_KEY'},
    json={
        'audio_url': 'https://example.com/source-audio.mp3',
        'voice': 'adam',
        'model_id': 'eleven_multilingual_sts_v2',
    }
)

audio_url = response.json()['audioUrl']

View full API documentation

Frequently Asked Questions

How is voice cloning different from text-to-speech?

Voice cloning (speech-to-speech) transforms existing audio into a different voice while preserving the original timing, emotion, and delivery. TTS generates speech from text.

What voices are available?

We offer built-in presets (adam, rachel, josh, bella, etc.) and support any ElevenLabs voice ID for custom voices.

What's the maximum audio length?

The API supports audio files up to 10 minutes. For longer content, we recommend splitting into chunks.

Is voice cloning ethical?

Vydra is designed for legitimate use cases like content creation and accessibility. Users are responsible for obtaining consent when cloning real voices.

Ready to Get Started?

Start Building Free Read the Docs