How to add an AI voiceover
- Move the playhead to where the voice should start.
- Type your script in Text to speech in the AI tab.
- Choose a voice and speed. Speed goes from 0.7× to 1.4×.
- Press Add voiceover at playhead. The voice appears in the Audio tab, where you can move it, change its volume or delete it. Other sound dips under it automatically if Lower other sound under voiceovers is ticked.
The voices
| Voice | Accent | Character |
|---|---|---|
| Heart | US | Warm female, a good default |
| Bella | US | Bright female, upbeat promos |
| Nicole | US | Soft, close — calm tutorials, ASMR-style |
| Michael | US | Clear male, explainers |
| Fenrir | US | Deeper male, trailers and drama |
| Emma | UK | Crisp female, documentary |
| Isabella | UK | Friendly female |
| George | UK | Measured male, narration |
| Fable | UK | Storyteller male |
Writing a script that sounds natural
- Write for the ear. Short sentences, one idea each. Read it aloud before generating it.
- Use punctuation to pace it. Commas add small pauses, full stops longer ones. Split long paragraphs into separate voiceovers so you can time each one to the picture.
- Spell out tricky words. Write "twenty twenty-six" or "V D O app" if the voice reads a number or name oddly.
Frequently asked questions
Which languages are available?
The voices speak English, with US and UK accents. For other languages, record your own voice in the Audio tab.
How big is the download?
The voice model is about 88 MB and is downloaded once. After that it works offline.
Can I use the voiceover commercially?
The Kokoro model is released under the Apache 2.0 licence, which permits commercial use. You're responsible for the script you write.
Can I caption the AI voiceover?
Yes. In Auto captions, tick the voiceover under Caption these sounds.