Automatically find every person in your footage, choose exactly who stays visible, and export a clean, permanently redacted video — without a single frame ever leaving your device.
Unless you want it to be — manual control is there when you need it.
Drop in any video. Formats your browser can't play natively are converted automatically before scanning starts.
A full pass runs before playback unlocks, finding every person in the footage using an in-browser detection model.
Everyone's blurred by default. Un-blur, lock, or hand-draw exactly who and what you want, then export an MP4.
Redactrr watches the entire clip once, using an in-browser AI model, before playback ever unlocks. That means every person is already found and blurred by the time you start reviewing — not discovered reactively as you scrub through.
Every detected person is blurred by default. Click them directly on the video, or flip their card in the sidebar, to make them visible instead. They keep being tracked automatically either way, no matter where they move.
For anything automatic detection doesn't catch — a partial view, an object, a sign — draw a region by hand. It's fully independent of the AI model, so it works even if detection fails to load entirely.
Once you're certain someone should stay blurred, lock it. A locked person can't be un-blurred by an accidental click on the video or the sidebar — and they're guaranteed to render blurred in your final export.
Exporting renders every redaction directly into the video's pixels. Watch it move through recording and MP4 conversion in the Jobs panel, then download whenever it finishes — nothing downloads automatically without you.
Switch themes any time. Every color, including text contrast, is tuned per theme rather than reused — so light mode is genuinely readable, not just the dark palette inverted.
Detection, blurring, and export all happen locally in your browser. There's no server involved in processing your video.