AI avatar generator for video ads

What is an AI avatar?

A generated person who can be filmed saying anything you write. For an ad that means a presenter who holds the product, looks at the camera and speaks, without a shoot day or a release form: User Generated Content, with the creator generated too.

The useful distinction is between an avatar that talks and an avatar that sells. A floating head on a background reads as software; a person in a room with your product reads as an ad.

How many presenters are there, and who are they?

42, two per language across 21 languages, each speaking natively rather than dubbed. None of them is a real person, and none is licensed from a real person.

You can also upload a photo of the face you want. That face becomes the presenter and speaks the script.

• Mary English
• James English
• Yinuo Mandarin Chinese
• Yichen Mandarin Chinese
• Sunita Hindi
• Ram Hindi
• Maria Spanish
• Jose Spanish

Does the voice match the face?

Each presenter speaks in their own voice, in their own language, and the lip sync is generated with the audio rather than fitted to it afterwards.

The words are said exactly as written. If you lock your copy, that is what comes out of their mouth.

Where does an AI avatar still lose to a real person?

A named face with an audience, and any claim that depends on someone genuinely having used the product for months. A generated presenter must never pretend to a personal history it does not have.

For everything else, which is most performance creative, the audience never asks and the ad account never notices.