Google Vids Now Turns Text and Selfies Into Finished Videos

Key points

• Google is adding two new tools to Google Vids: Gemini Omni for generating and editing clips from text, and personal avatars that mimic your face and voice.
• Personal avatars need only a selfie and a short voice recording, and users can type what they want the avatar to say.
• Both features are limited to paying Google AI Pro, Ultra and Google Workspace business subscribers, with avatars restricted to users 18 and over in certain regions.
• Every clip carries an invisible SynthID watermark so viewers can check whether a video was made by AI.
• The rollout builds on Veo 3.1, Google's video generation model, which the company opened to all Vids users in February.

Google just made it a lot easier to fake yourself on camera. And that is not entirely a joke.

The company has rolled out two new features inside Google Vids, its video creation app, that together let you generate a clip from a written sentence and stick a digital version of your own face into it. The update was announced by Google AI.

Here is the short version.

What can the new Google Vids actually do?

It can build a video from a few sentences, and it can put a lookalike of you on screen without a camera. The first tool is called Gemini Omni. The second is called personal avatars.

Gemini Omni is powered by Google's video generation system, the software that creates moving pictures from written instructions. You type what you want to see, in ordinary language, and it produces a clip. You can also drop in a photo or a rough sketch as a reference, and Omni will try to match the style or the subject.

The editing bit is where it gets interesting for anyone who has ever wrestled with video software. Instead of dragging sliders, you chat. You can ask it to swap a background, brighten the lighting, or add an effect, and it will change just that part without you starting over. This works on clips you generated with Omni and on ordinary videos you filmed on your phone.

Then there are personal avatars. Upload a selfie and a short voice recording, and Vids will build a digital stand-in that looks and sounds like you. Type a message, and your avatar delivers it to camera. No hair fixing, no lighting rig, no take three.

Who gets to use it, and what does it mean for the rest of us?

Access is paid and limited. The features are available to subscribers of Google AI Pro and Ultra, the company's consumer AI plans, and to Google Workspace business customers, which is the paid version of Gmail, Docs and Drive that companies buy for their staff. Personal avatars are further restricted to users aged 18 or older in certain regions, and Google says each avatar is tied to the account holder's own likeness.

That last point matters. One of the big worries with a tool like this is deepfakes, videos generated by AI that convincingly show a real person saying things they never said. Google's answer is twofold. Avatars are locked to your own face and voice, on your own account. And every generated clip carries a SynthID watermark, an invisible marker baked into the pixels that lets software later confirm the video was made by AI.

A watermark you cannot see is not a magic shield. It only works if the platform showing the video knows how to look for it. But it is a real, checkable signal, which is more than most consumer video generators offer today.

Is this fun, or just efficient?

Honestly, both. Sending a personalised video message to a colleague without brushing your hair is a genuinely useful trick. So is fixing bad lighting by typing "fix the lighting".

The flip side is the one every creator will feel. When anyone can produce a slick 20 second clip in the time it takes to write a text, the value of any single slick clip drops. The interesting videos will be the ones with a point of view a machine cannot guess.

For now, if you pay for Google's top AI tier, you have a new toy on Monday morning. Everyone else gets to watch what people make with it.