OpenScreen

Each take is automatically zoomed where you clicked, by a fixed rule — no AI, no network call — and can be turned off in Record mode. Zooms, speed ramps, trims and comments appear as pills on the timeline.

New recordings open with a wallpaper (image, color or gradient), padding, shadow, rounded corners and light motion blur already applied; every effect composites live and can be changed without a re-shoot.

The pointer is recorded apart from the pixels, so its size, smoothing, motion blur and click bounce can be restyled after the take, from five cursor packs (Studio Ink, Prism Glow, Pop Coral, Pixel Candy, Star Sprout) that can be drawn as 3D objects with a shadow.

Zooms can stay flat, turn the screen left or right, or orbit it while following the zoom focus, with depth of field; the take can be placed in a window, laptop, phone or screen frame over an animated Drift, Aurora or Waves background.

One take with a webcam can be picture in picture, side by side, stacked or full screen; a local segmentation model can blur, cut out or replace the room behind you with no green screen and nothing uploaded. A Full Camera mode grows the camera to full screen for a stretch of the timeline.

Transcription runs on your machine via whisper.cpp (model downloaded once on first run, then works offline; audio never leaves the laptop). Delete a word or a silence and the cut lands on the timeline non-destructively. Captions can be styled (typeface, size, color, position) and burned into the ren

Reads the actual transcript and timeline to propose cuts with checkable timecodes, reporting how much time each edit saves. Every edit is an ordinary undoable one; it is off by default and needs a provider key you supply.

Records through the OS capture path: ScreenCaptureKit on macOS, Windows Graphics Capture on Windows (with system audio out of the box), and PipeWire via the ScreenCast portal on Linux.