Raw footage in.
Edited video out.
Cuts the dead air, subtitles what people say, punches in on the big moments and adds sound effects. Then tell it what to change. Everything runs on this device — nothing uploads.
Cuts the dead air, subtitles what people say, punches in on the big moments and adds sound effects. Then tell it what to change. Everything runs on this device — nothing uploads.
Working…
Type a change in the box under the timeline, or tap one of these.
Speech is transcribed on this device. The first time, a 40 MB speech model downloads and is kept for next time.
A whoosh on each change of shot and a hit on each zoom.
Tap to add a layer to this clip. Tap the highlighted layer again to remove it.
Drag text on the preview to move it.
Safe zones mark where Instagram and TikTok put their own buttons. They never export.
Encoding… 0%
Everything works without this. When a change you type isn't understood, a key lets an AI model work out what you meant — an Anthropic key (starts sk-ant-) uses Claude, a Google AI Studio key uses Gemini. The key stays in this browser and is only sent to that provider, with the words you typed.