Topic → script → footage → voice → captions.

Edit the title, script, or scene footage queries before footage fetch begins.

Clone your own voice once so every video uses it instead of the shared stock niche voice. Requires a real audio sample of the voice being used — ElevenLabs' cloning terms require you have rights to it.

Image or video layered on top of the footage — logo bug, lower-third, callout graphic. Composited at export/recompose.

Text rendered on top of the footage — lower-third, stat callout, title text. Composited at export/recompose.

A time-ranged video filter — vignette, grain, blur, grayscale, or darken. Applied directly to the footage at export/recompose.

Camera/zoom motion applied to the selected clip's footage. Real effect, applied by pipeline/motion.py at render time.

The cut/transition used going INTO the selected clip from the previous one. Resolved to a real FFmpeg xfade filter at concat time.

Click words to cut (hide) or restore them. Cuts are non-destructive ops — original transcript unchanged.

Search for new footage to replace the selected clip. Results pulled from Pexels/Pixabay.

Change the on-screen caption/key phrase for the selected scene.

Select a new background music track from the asset library.