Once your custom avatar is saved to My avatar, you can make it speak by adding text and selecting a voice (or using the audio from its reference video), then generating the video. Compatible models—including the Kling 2.6 and Kling 3 series in audio mode—support lip-sync.
Before you begin: Add an avatar from the Public avatar or My avatar tab to your current project.
Choose how your avatar speaks
Use a Voice from the library
Enter your script, then choose a Voice that matches the language and style of your video.
Enter the text you want the avatar to say.
In the Creation bar, select Add (+), then select Voice.
In the Select voice panel, open the Auto language menu and choose the language. To browse Voices from all supported languages, leave Auto language selected.
Choose a Voice from the filtered list.
Tip: Use punctuation to create natural pauses. Keep sentences short and clear.
Use your own voice
To use your own voice, upload a video with a clear recording of your voice when you create custom avatar. Medeo will use that audio as the avatar's voice.
For video requirements and upload instructions, see Create a Custom Avatar.
Generate and review the video
Select Generate. Once the video is ready, check that the speech and mouth movements are in sync.
Fix speech and lip-sync issues
No sound: Make sure that the avatar and the selected voice or recording have been added to the current project.
The mouth movements do not match the speech: Shorten long sentences or silences. If you used a recording, replace it with one that has no background noise, distortion, or overlapping voices.
If the issue continues, use a short text or recording to generate a test video.