Cuescribe

Horizontal to vertical, following the speaker

Drop a 16:9 interview, podcast or vlog. Faces are found on your device and the 9:16 crop follows the speaker like a calm camera operator: it ignores small moves and glides when they walk. Or keep the whole frame over a blurred copy. Export 1080×1920 with the original sound.

How to use it

  1. Drop the horizontal videoInterviews and talking heads work best.
  2. Choose how to reframeFollow the speaker's face, a fixed crop, or the whole frame on a blurred background.
  3. Pick the shape9:16, 1:1 or 4:5.
  4. ExportAn MP4 at the new shape, with its sound.

What you can download here

  • Vertical or square MP4crop path smoothed so it doesn't jitter
Free Chrome extension

Cuescribe for Chrome

A transcript side panel on every YouTube and TikTok video, a library of everything you transcribe, and one click to open any of it in these tools.

Get the extension
Pro · early access

Pro keeps this free

Every tool stays free. Pro (early access, $9/month) adds batch delivery for client work and the extension's Pro features; longer files and multi-file projects are in development. Early subscribers keep their price and choose what gets built next.

What Pro includes

Next steps

Add captions to videoStyled, burned-in captions, 9:16 crop and safe-area guides.Clip from transcriptClick sentences, export that clip as MP4 with its subtitles.Video QC checker NewBlack/flash frames, freezes, bars, clipped or missing audio, loudness.Filler word removerUm, uh, false starts and repeats cut; keeps your last take.Silence removerCut dead air; get the tightened video, an EDL or a WAV.Apply a LUT New.cube LUT on video or photos, before/after, export.

All 44 free tools →

Questions

How does it decide where to crop?

It looks for faces five times a second and follows the largest one. Small movements are ignored so the frame stays steady, and one-frame mistakes are filtered out. You see the crop on the source as it plays, before you export.

What about two people talking?

It follows the larger face (usually the one closer to the camera). For a two-shot where both matter, use "Whole frame, blurred background", or a fixed crop.

Is anything uploaded?

No. The face detector (Google's open MediaPipe model) downloads once and runs in this tab, and the video is encoded on your computer.

Can I add captions too?

Yes. Export the vertical video, then open it in Add captions to video to burn in word-by-word captions.