How to add captions to a video

Last updated 2026-10-20

To add captions to a video, run auto captions to transcribe the speech, then read every line and fix names and jargon. Pick a style that reads on a phone: large text, two lines at most, clear of the platform's buttons. Burned-in captions show everywhere, because they are part of the picture. Woven burns them in.

Four of Woven's caption looks on the same clip, exported from Woven. Pick one to compare.

Auto captions or manual captions

Auto captions transcribe the speech for you. Manual captions mean typing every line and setting its timing. Run auto first, then edit what it got wrong. Type from scratch only when the clip is a few seconds long, or when the words are not spoken.

Platforms offer their own auto captions. YouTube says its automatic captions come from machine learning, so quality varies, and it asks you to review and fix them. Instagram turns a reel's speech into text at the bottom of the screen with speech recognition, controlled by a toggle named Enable closed captions. TikTok Studio lists Auto captions to add and edit captions for your posts. Those are platform captions, and they live on that platform.

Check names, numbers and jargon

Machine captions fail on the words that matter most. YouTube says its captions can misrepresent speech because of mispronunciations, accents, dialects or background noise. Expect the damage in names, numbers and product terms.

Do one pass for those words. Compare every name, number and term of art with what you actually said. A wrong brand name is worse than a missing filler word.

If you switch languages mid-video, read that part line by line. YouTube says speech in several languages at once can stop automatic captions from being made at all. Woven captions one language per run, so a reel that mixes languages may caption the second one poorly. Fix the text before you choose the look.

Make captions read on a phone

Test on a real phone, not only on your Mac screen.

  • Size. Large enough to read at arm's length.
  • Lines. Two lines at most. Short lines are quicker to take in than a wrapped paragraph.
  • Position. Keep captions above the platform's own buttons and text. Instagram's closed captions sit at the bottom of a reel, so burned-in captions in the same spot can stack on top of them. For ads, TikTok says the safe zone changes with the length of the ad's text caption, so check its safe zone files when your text runs long.
  • Contrast. W3C's web accessibility guideline asks for 4.5:1 between text and background, or 3:1 for large text. Treat that as a minimum, and add an outline or shadow over bright or busy footage.

Burned-in or platform captions

Burned-in captions are part of the picture, so they look the same on every platform and viewers cannot turn them off. Platform captions are made and controlled by the platform. The open versus closed captions guide compares the two in depth.

Woven only burns captions in. It cannot export a caption file such as SRT.

Animated captions and the five looks

Animated captions highlight or reveal words as they are spoken. Match the look to the speech. Fast talking suits a highlight that guides the eye. Slow or technical talking suits plain subtitles, because motion on every word gets tiring.

The five caption looks in Woven
LookWhat showsSuits
HighlightThe whole caption, spoken word emphasizedFast speech
SubtitleThe whole caption, no emphasisCalm or technical speech
Word by wordOne centred word at a timeOne-line hooks
Word in placeOnly the spoken word, where it sits in the layoutDesigned layouts
Progressive revealEach word from when it is spokenLines that build up

Add captions in Woven

Woven transcribes the speech in your clips, voice and audio, then burns the captions into the exported MP4. Transcription uses credits; the pricing page has the rate. Captions are fixed in the video, so choose the look and the position before you export.

Ask in chat

  • Add captions and use the Word by word look.
  • Change 'Jason' to 'JSON' in the captions.
  • Raise the captions above the bottom fifth of the frame.
  • Narrow the caption width so it wraps to two lines.

Or do it by hand

  1. 1Transcribe. Choose Generate captions from the Add menu. Woven captions the speech in your video, voice and audio clips; music counts only when nothing else is spoken.
  2. 2Match the look to the speech. With the captions selected, go to Appearance on the Style tab and open Caption style. Try Highlight for fast talk and Subtitle for slow talk. Preset offers None and Kinetic.
  3. 3Place and size. Drag the captions on the canvas so they clear the bottom of the frame. On the Edit tab, set Font, Weight and Size.
  4. 4Fix words. Ask in chat to change a wrong word. The caption keeps its timing and styling. To split or join a caption, right-click it.
  5. 5Check and export. Watch the preview once with the sound off, then export.

Common questions

Sources

Try Woven free for 3 days

Then $25/mo billed yearly ($299/yr) or $39/month. Native Mac app for short-form video.

Download for Mac

Paste into Claude Code to install and connect Woven.

Related