Get suggestions grounded in speech
AI reads the transcript and suggests self-contained excerpts with timestamps and an explanation. Sources can be up to three minutes.
Find the idea. Review the cut.
Get transcript-based clip suggestions with reasons. Review the moments you want, then export selected clips separately or download them as a ZIP.
Transcript-based suggestions · You approve the clips
Up to 3 minutes per source; no automatic virality guarantees.
Up to 180 seconds and 100 MB per source. Sign in to process and export. Importing may be blocked by the source platform; file upload is also available.

Workflow illustration using our sample footage.
The result joins seconds 0–4 and 8–11 from the source into a seven-second video.
These are real exports of our self-made sample, with synthetic audio and hand-written example captions. Upload a video with speech to try automatic transcription and translation.
AI reads the transcript and suggests self-contained excerpts with timestamps and an explanation. Sources can be up to three minutes.
Jump to each suggested moment, read the reason and select the clips you want. You make the final editorial decision.
Render your selected clips as individual MP4 files and download them together in a ZIP. Their combined length must stay within 180 seconds.
It uses the video transcript to recommend excerpts and explain each choice. It does not analyze visual action or guarantee views. Review each suggestion before exporting.
Not yet. Each source must be 180 seconds or less and under 100 MB. Prepare a short excerpt from a longer recording first.
Yes. Use the manual clip controls to set your own start and end times. To export the original recommendations separately, select them in the suggestions list.
Use footage you own or have permission to edit. Only upload material that can be public: the current editor uses a public asset bucket.