Vertical video
A talking-to-camera video, reframed vertical and captioned
Give No Prod AI one recording of somebody talking to camera. It finds the face, frames a 9:16 picture around it, writes down what was said, and captions it in short bold lines.
Free for 14 days, no card needed.
How it goes
- 1
The face found
About once a second through the recording, so a steady framing is chosen rather than one that chases every movement.
- 2
Framed to fill the screen
Eyes on the upper third, cropped to fill a vertical frame — never closer than the footage stays sharp.
- 3
Captioned
Transcribed on your computer in the language spoken, and cut into short bold caption lines.
- 4
Rendered
A 720×1280 video of up to 90 seconds, sized to send.
Change anything
The result is an ordinary edit: move the framing, restyle the captions, trim it.
Small enough to send
Rendered at a capped size — under about 41 MB for 90 seconds — so it travels by message or upload.
Nothing uploaded
The face is read and the video made on your machine.
Questions
What if no face is found?
It frames the middle of the picture and says so, and you can move the framing yourself in the editor.
How long can the video be?
This pipeline keeps the first 90 seconds, the length a vertical video is usually watched at. For a longer recording, the editor and the other pipelines have no such limit.