Guide
How Editz actually 'sees' your footage
It's not just reading a transcript — here's what's really happening.
A lot of "AI video editors" are really just auto-captioners wearing a trench coat — they transcribe your audio and cut on silence. That's useful, but it's blind to anything visual: composition, motion, which take actually has the good take.
Editz's approach layers a few real signals instead of one: a local Whisper transcript for word-level timing and content, beat detection for anything music-driven, and a shared timeline document that every tool reads and writes to — so a cut made by one tool call is immediately visible to the next one, and to you, in the live preview.
Why this matters in practice
It means you can chain instructions naturally — "cut the filler, then sync it to the beat, then grade it warm" — and each step builds on the last, instead of you re-exporting and re-uploading between every change.