AI Avatar
Create a talking video from saved photos and a custom voice.
An AI avatar combines a saved photo with a custom voice to create a talking video. Save the avatar once, then use it for new scripts.
Create an AI avatar
Prepare a ready custom voice before you start.
- Open video generation in your project and select AI Avatar.
- Select New AI avatar or Create AI avatar.
- Enter a name and select the custom voice.
- Add one to five clear photos.
- Confirm that you can use the person's image and voice.
- Select Save AI avatar.
Each photo can be up to 10 MB. Source photos stay private to the workspace. All workspace members can use the avatar.
Generate a video
- Choose your AI avatar and the photo to use for this video.
- In Create voiceover, write your script and create the audio. You can also choose a matching recording from Saved voiceovers.
- Listen to the final read. The voiceover must be 4–30 seconds long.
- Select Approve voiceover.
- Select Generate AI avatar video.
AI avatar video generation requires an active paid plan. Each generation uses 1 video credit. Generation continues in the background after the request starts. You can use the generated video in your project.
If you change the voiceover speed, save the new speed version before you approve it. A voiceover outside the duration limit remains in Assets, but cannot be used for an avatar video.
Manage an avatar
Use Manage AI avatar to rename an avatar or manage its photos. Disabling an avatar removes its source photos. Existing generated videos stay unchanged.