Video Summary turns a long recording into a short one that keeps only what was actually said. Every kept moment holds on to its original timestamp, so one tap takes you back to that exact second in the full video.
14-day free trial, then $6.99/month. Everything runs on your phone.
Record in the app or import a video you already have. Choose how long the digest should be. Watch the app find the parts worth keeping.
Five, ten, twenty or thirty percent, with a live estimate. Change your mind later and it re-cuts in seconds.
The waveform lights up section by section as the app picks them. No spinner, no mystery.
Tap any timestamp to open the full recording at that exact second, with the kept stretch shaded on the scrubber.
Lectures, interviews, meetings, webinars, and your own raw footage. Videos you own, on a phone you own.
The phone's own speech recognition supplies the words. Scoring is classical — term weighting, topic segmentation, cue phrases, prosody — not a guess from a chatbot.
Nothing is sent anywhere. No video, no audio, no transcript. The analysis runs on the device, in the dark, on a plane, wherever you are.
Video Summary only ever reads your file. The digest is a new video beside it, and deleting a digest never touches the recording it came from.
Search the words inside your digests, not just their names. Rename, re-cut at a different length, or delete — your digests stay yours.
Five screens carry the whole product. This is all of them.
This is an architectural fact, not a policy promise. There is no upload path in the app.
It depends on the phone and the recording. The slow part is reading the audio and recognising the speech; an hour-long video is minutes, not seconds. The screen shows you where it is the whole way through, and you can cancel at any point.
The app falls back to audio patterns — energy, pauses, emphasis — and says so on the digest, so you always know how it was made.
Yes. The ranking is kept with the digest, so a different length re-cuts without analysing anything again.
Every digest you have already made stays playable, forever. The subscription covers making new ones.
Yes, apart from the speech recognition on devices that need a language pack downloaded first. Nothing about the analysis requires a network.