Choose and update an AI voice
Choose a voice during setup, or open the voiceover controls in Audio.
Search or filter the available voices by language and other offered attributes. Listen to a sample.
Apply the voice and generate the narration.
After editing the script, select Apply Changes and choose the offered scope for changed scenes or the full script.
Wait for generation, then check that spoken words and captions match.
Regenerating the full voiceover creates new audio for the entire narration. Review the confirmation and account-usage information. A project using an uploaded recording plays the original speech instead of synthesizing the script.
Correct pronunciation and pauses
From the story menu, select the word or phrase and open the pronunciation control. Enter the offered pronunciation guidance and listen to the sample. Apply it to the relevant scene or use the all-scenes option for a repeated term. Use the pause control at the narration cursor when you need a pause. Apply voiceover changes, then listen to the full sentence. These controls cannot alter words already spoken in an uploaded recording.
Add music and balance audio
In Audio, browse or search music, listen to a sample, and apply the track. Use the removal control to take it off the project. Open audio settings to adjust Background music and Voiceover. Keep speech clear. Music-volume changes can be previewed directly. If a voiceover-volume change marks narration as out of date, use Apply Changes.
Check scene controls for source-clip sound, narration, and music. Review the scope before applying mute to all scenes. Listen to the intro, ending, and transitions for competing audio.
Use an AI avatar
Select a scene and open Avatars. Choose a stock or custom presenter, apply it, and follow the voice and generation prompts. Avatars can be previewed with their lips blurred. The final lipsync happens at video generation time.
Select the avatar to adjust the available position, size, crop shape, background, and border controls. Keep it clear of captions and important content. Use the visibility control to show or hide the avatar. Use the all-scenes controls when you want the same change throughout the video.
Create a personal avatar
In Avatars, choose the custom-avatar creation option and read the instructions.
Upload or record clear, unedited training footage of one person speaking naturally to the camera. Supported uploads include MP4, WEBM, and MOV.
Review the details and name the avatar.
Record or upload the consent statement displayed in the wizard and complete the consent step.
Submit, wait for processing, and apply the completed avatar from your library.
Follow the duration, file-size, and recording requirements shown in the wizard. Use a steady camera, clear speech, good lighting, and little background noise.
Create a voice clone
In the voiceover controls, find your voice-clone section and select Create new.
Read the instructions and upload or record clean training audio. Sample uploads support WAV, M4A, and MP3.
Review the sample information and name the clone.
Complete the consent recording using the displayed statement and submit.
Choose the completed clone from the voiceover library, generate narration, and listen to the result.
A clone speaks new script text with an AI voice. An uploaded recording plays the original take. If avatar or clone generation fails, check the displayed requirements and retry with corrected training or consent material.