Google Vids adds prompt controls to direct AI avatars

In short: Google is adding a way to “direct” AI avatars in Google Vids by typing prompts, instead of only choosing from preset characters.

What happened

Google Vids is a video editor inside Google Workspace, which is the paid set of tools that includes Gmail, Docs, and Drive for work and school. It already lets people add an AI avatar, which is a digital on-screen presenter that speaks your script with matching mouth movements.

Until now, Vids mainly worked like a menu. You picked from preset avatars in different styles, like photorealistic, 2D cartoon, or 3D cartoon, and then pasted in a script for the avatar to read. Users could preview voices, generate short clips, and drop them into a scene, but they could not really change how an avatar looked or behaved beyond the preset choices.

TechCrunch reports that Google is now adding a way to customize and instruct these avatars using prompts. A prompt is just a typed instruction, like telling the system what you want in plain words (similar to giving directions to a contractor and describing the result you want).

Why it matters

For regular workplace videos, like training, updates, or simple explainers, this could reduce the need to record a person on camera. It may also make the videos feel more “on purpose,” since teams can ask for a specific tone or style instead of settling for whatever the preset avatar happens to do.

Source: TechCrunch AI