One voice per character
A character in the studio now has a voice of its own. You pick it once in the character folder, after listening to the options, and from then on every clip with that character can use it.
Before this, a talking clip meant recording audio, uploading it and syncing a mouth to it. That still works, but for most social clips it is more than you need.
Write the line, get the clip
Choose Kling 4.0 with your character, describe where the character is and what happens, and put the spoken line in quotes. The model makes a clip where the character says that line, with the face and the voice you saved.
Keep a line short. One or two sentences fit in a clip; a paragraph does not.
- Describe the place and the action, not the face: the face comes from the character.
- Put the line in quotes so the model knows what is spoken.
- Choose 9:16 for Reels, Shorts and TikTok.
Why we chose Kling 4.0 for this
Before we built it, we compared models on the same characters and the same lines, without knowing which clip came from which model. Kling 4.0 with the character's own voice gave clips where the voice and the face belonged together. That is the one we put in the studio.
Questions
Can I still use my own recording?
Yes. Upload your audio and use a lip sync model or a talking photo. The saved voice is the quicker route for short lines.
Can an AI assistant do this for me?
Yes. Through the connection an assistant can set a voice on a character and make the clip, if you approve it.
Read next
Try it in the studio
Open the studio and try what you just read. You start with 1,000 free credits.
Get early access



