1. Provide Video
Drag & drop videos — or a whole folder — here
or
One URL per line to use several clips.
The folder has to be shared as Anyone with the link. Every video in it is used, in name order.
2. Styling
Each clip leads with a different angle. A strong moment can appear in more than one. Clips render one after another, so this takes roughly as long as that many single videos.
Transcribes everything, then drops false starts, rambling and weaker duplicate takes — and reorders when a later line makes a stronger opening. You can adjust the cut afterwards.
AI writes it from your script and decides what to ask for — follow, save, DM, book a call. Leave the field empty to let it choose.
Picks the strongest frame from the video and puts your title on it, styled like the captions.
Cuts pauses longer than 0.4s (and dead air at the start/end), keeping a short breath.
Cuts "öh"/"hmm"-type sounds. Still being tuned — it can occasionally clip real speech.
Generated for this cut — mood from what you said, structure from your timeline. Music is never subtitled.
Narrates footage with nobody speaking in it — photos, b-roll, screen recordings. The narration is subtitled too. Skipped automatically if someone is already talking.
Applies to photos. A square product shot is laid on the canvas over a blurred copy of itself rather than cropped, so nothing is cut off. Video keeps its own shape.
Processing Video
Uploading and starting AI transcription...
Done
5. Edit Transcription
Click any word to fix spelling or capitalization. Do not add or remove words, or the timing will break.