AI Stunt Double. Video, done for you
1 face, 1 voice, any setting. Nobody filmed.
The AI Stunt Double page on this site is its own case study. The man on it does not exist. We made 1 face, placed it in 9 settings, gave it a voice, and made it talk.
That is the service: we clone a client once, their face and their voice, and produce 20 videos a month without them recording. An editor reviews every video.
How it works
- InA short recordingA few minutes of the client on camera, plus the ideas they want to talk about.
- AIThe doubleA face model and a voice model. Scripts become speech, speech becomes video, in any setting.
- PersonThe editorChecks every video for the face, the words and the facts before it goes out.
- Out20 videos a monthShort videos, posts and a newsletter, on the client's channels.
What AI does
- Keeps 1 face the same across every scene: a studio, a run, a stage, a kitchen.
- Turns a written script into speech in the cloned voice.
- Fits the mouth to every word of the speech.
- Puts the words on screen so the video reads with the sound off.
What people do
- Write the scripts from the client's ideas, and the client approves them.
- Pick the settings and the shots.
- Review every video before it is published.
- Make the videos of a person only with that person's permission.
By the numbers
- 1
- face, kept the same in every scene
- 9
- settings on the product page
- 2
- talking clips, wide and upright, 1 voice
- 20
- videos a month on the Monthly plan
What is under the hood
The face was made first as a portrait, then placed in each setting with the portrait as the reference.
The voice is a library voice from ElevenLabs. The talking clips were made from a still and the voice file, then a lip sync pass fitted the mouth to the words.
Every picture and clip on the page carries a Made with AI mark, and the page says twice that the person does not exist.
Want a system like this for your business?
Tell us what the work is. We reply with what we would build and what running it costs.
