Voice Cloning Guide

Voice cloning lets you replicate a specific voice from a short audio recording and use it for character dialogue and narration in your projects.

What you need

• A paid plan (voice cloning is not available on the free tier)
• A clear audio recording of the voice you want to clone (10 to 30 seconds)
• Permission to use the voice (more on this below)

Preparing your reference audio

The quality of your cloned voice depends heavily on the quality of your reference recording. Follow these guidelines for the best results:

• Record in a quiet environment. Background noise, echo, and room reverb will be captured in the clone.
• Use clear, natural speech. The person should speak at a normal pace and volume, as if having a conversation.
• Keep it consistent. Avoid dramatic changes in tone, volume, or pacing within the reference clip.
• Aim for 15 to 20 seconds. This is the sweet spot. Shorter clips may not capture enough of the voice's character. Longer clips don't necessarily improve quality.
• Avoid music or sound effects in the background of the recording.

Supported formats

• WAV, MP3, FLAC, or OGG
• Maximum file size: 20 MB

Step 1: Open your Media Library

Navigate to your Media Library from the project workspace. This is where all your uploaded audio files and voice references are managed.

Step 2: Upload your audio file

Click Upload and select your reference audio file. Choose Voice Reference as the upload type so the system knows this is for voice cloning.

Step 3: Confirm voice consent

Before the clone is created, you'll be asked to confirm that you have permission to use this voice. This is a legal requirement.

You must confirm that one of the following is true:

• It's your own voice.
• You have explicit permission from the person whose voice it is.
• The voice is from a public domain or licensed source.

Step 4: Use the cloned voice in your projects

Once your voice reference is uploaded and consent is confirmed, the cloned voice appears in your voice picker. You can assign it to any character in your script, just like a preset voice.

Select the cloned voice from the voice picker when configuring a character, and any voice segments for that character will use your cloned voice.

Tips for great cloned voices

• 15 to 20 seconds of clean speech is ideal. More audio doesn't always mean better results.
• Diverse sentence types help. A reference that includes both statements and questions gives the AI more vocal range to work with.
• Test before committing. Generate a short test segment with the cloned voice before using it across an entire project.
• Try different reference clips. If the first result doesn't sound right, try uploading a different recording of the same person.

Managing your cloned voices

You can find all your cloned voices in your Media Library. From there you can:

• Preview the reference audio
• Rename the voice for easier identification
• Delete the voice reference, which also withdraws your consent for that voice

Deleting a voice reference removes it from the voice picker. Any previously generated audio using that voice will remain in your projects, but you won't be able to generate new segments with it.