Plan Requirements
Custom voice cloning is available on Pro ($30/mo) and Creator ($50/mo) plans. Plus and Free plans can use built-in voices only.
Step-by-Step Setup
- Go to Character Settings → Voice → Clone Voice
- Upload an audio file in MP3, WAV, OGG, or WEBM format
- Wait for processing: typically 2-5 minutes depending on file length
- Preview the cloned voice with a test phrase
- Save to apply the voice to your character
Audio Sample Guidelines
The quality of your clone depends almost entirely on the quality of your source audio:
- Duration: minimum 30 seconds, but 60-120 seconds produces significantly better results
- Clarity: clear speech with minimal background noise and no echo
- Consistency: speak at a natural, even pace throughout the sample. Avoid sudden volume changes
- Variety: use different sentences rather than repeating the same phrase. This gives the model more vocal data to work with
- Format: lossless formats like WAV produce better results than compressed formats like low-bitrate MP3
Recording Tips
- Record in a quiet room: a closet with clothes works surprisingly well for sound dampening
- Use a real microphone if possible, even a basic USB condenser mic
- Keep your mouth 6-12 inches from the mic
- Read a paragraph from a book or article to get natural, varied speech
Replacing a Voice Clone
You can upload a new audio sample at any time to replace the existing clone. Go to Character Settings → Voice → Replace Clone. The old voice is deleted immediately and the new one takes effect after processing.