AI Clip Generator

Recapo transcribes the speech, tracks scene changes, and weighs emotional peaks against dead air. It then ranks candidate segments according to the selected goal: best hook, talking clip, or 9:16 short. The ranking reduces scrubbing time, but users still review, reorder, trim, reject, or add segments before export.

Yes. The ranked list is a starting point, not a verdict. You can reorder candidates, fine-tune every in and out point, drop clips entirely, or add a stretch the AI did not flag.

Anything with speech or clear scene structure: films and shows for recap highlights, livestream replays, interviews, lectures, and gameplay. Speech-light footage still works, but ranking leans more on visual scene changes.

Recapo does not publish a fixed number of clips per source. The ranked list depends on video length, signal density, and the chosen clip goal. Review the candidates and export only the keepers; the public page currently describes one MP4 export per task, so do not promise a batch count that the interface does not guarantee.

Choose the talking-clip goal when complete statements matter, because it favors quotable sentences with clean start and end points. Then read the transcript immediately before and after each proposed cut. Extend the in or out point when the speaker, setup, response, or conclusion is missing, and reject clips that remain ambiguous without the original video.

The public page does not currently document a dedicated search filter for a named topic or speaker. Use the closest clip goal and task instructions, then review the transcript and ranked candidates manually. If product search is added, update this answer only after the control is visible in the initial interface and has been tested on multi-speaker footage.

The selected clip can continue into editable captions, SRT or VTT export, subtitle burn-in, translation, AI voiceover, vertical crop, and cover layout inside the same project without re-uploading. Those are connected next steps; do not imply that every treatment is applied automatically when the initial clip candidates are generated.

AI Clip Generator is designed to analyze a long source, score moments, and return a ranked shortlist. AI Clip Maker emphasizes combining AI suggestions with manual precision when the user already knows which line or interval matters. Start with the Generator for discovery and use the Maker when deliberate manual selection is the primary task.