Guides
Voice Cloning
Clone any voice from an audio sample and use it for narration in your videos. Supports multiple providers for different use cases.
Choosing a Provider
ElevenLabs
Best overall quality. 30+ languages. Recommended for production.
Hume AI
Expressive with emotion control. Great for storytelling. English only.
Chatterbox
Open-source, budget-friendly. English only. Good for testing.
Audio Sample Requirements
- Duration: Minimum 30 seconds, 1-2 minutes ideal
- Quality: Clear audio, minimal background noise
- Content: Natural speech, varied intonation
- Format: MP3, WAV, M4A, or WebM
- Single speaker: Only one voice in the recording
For best results, record in a quiet room and speak naturally. Avoid reading in a monotone voice.
Recording Tips
- Use a good quality microphone (phone recordings work too)
- Speak in your natural voice - don't try to "perform"
- Include a variety of sentences and emotions
- Avoid ums, ahs, and long pauses
- Record in a quiet environment with minimal echo