Turn a long video into a vertical captioned Short

Cut the part you want, let speech recognition caption it, fix any misheard words, then crop the video to 9:16 with the speaker kept in frame and the captions burned in word by word. The crop and the captions are drawn in the same pass, so the video is encoded once, at 1080 × 1920, with the original sound.

Every step runs in your browser; your files are not uploaded.

Steps

  1. Cut the part you want (optional)

    A Short can be up to 3 minutes. Set the start and end with the player, convert and download. Skip this step to use the whole video.

    Open as a full tool: Video Trimmer

  2. Caption the speech

    Whisper runs on your device and writes timed captions from the speech. Pick the language if it is not detected, then send the captions and the video on.

    Open as a full tool: Auto Subtitle Generator

  3. Fix the words (optional)

    Play the video and correct names and misheard words. The word timings from the previous step are kept for every caption whose word count you leave the same.

    Open as a full tool: Subtitle Editor

  4. Follow the speaker (optional)

    The video is checked five times a second for faces. The largest face is followed, and the frame only switches to another person after they have been the main face for a second. The crop holds still while the speaker stays near the middle, glides when they move away and cuts cleanly when the picture changes.

    Workflow step

  5. Make the Short

    Choose the caption style (bold word by word, a black box, or an outline), how many words show at a time, the size, the height and the highlight colour. Keep captions above the bottom fifth, where the apps put their buttons. Then make the Short and download the MP4.

    Workflow step

FAQ

Is there a watermark?

No. The video is yours, with no watermark and no account.

How long can the video be?

Any length works for the captions and the crop, but a Short is up to 3 minutes. On phones, keep clips under about 3 minutes, because the whole video is held in memory while it is encoded.

What if there is no face in the video?

Skip the framing step and the crop stays in the centre, or choose a fixed position, or keep the whole frame with blurred bars above and below.

Made by BufferMint. We also design and build software products for founders. See our services →

Runs in your browser. Your files are not uploaded. We count anonymous visits, without cookies.