Talking avatar
How to make a portrait talk to your track: photo, audio, length and price.
The “Avatars” tab → Kling Avatar 2.0, OmniHuman 1.5 or VEED Fabric. The model syncs the lips and facial expressions of the person in the photo with your audio track.
- Add a face photo to the photo field.
- In the track field, choose a voice-over or a song from your generations. Create it first in the “Audio & voice” tab — you can’t upload your own audio file here.
- Optionally add a hint for the model — emotion, gestures, camera movement.
- Run the generation.
Length and price
The video is as long as the track: 2–60 s for Kling Avatar 2.0, 1–30 s for OmniHuman 1.5. The price is per second, so a longer track costs more; the exact amount is on the button.
Avatar videos are not part of the trial — they unlock after your first top-up. The photo must show you or a person who has given consent.