A lot of short video is watched with the sound off, at least for the first few seconds. If you talk to the camera, captions are the part of the edit that decides whether anyone stays long enough to turn the sound on.
An iPhone does not caption a video for you out of the box. This guide covers what the built-in apps can and cannot do, how auto-captions work, how to correct them, and the choice at the end: captions drawn into the picture, or a separate .srt file.
Key takeaways
- Photos cannot put captions on a video. iMovie can, one typed title at a time.
- Auto-captions are a draft. Names, jargon, and mixed-language speech are where a draft goes wrong, so plan a correction pass.
- Edit the video first and caption it second. Cutting a clip after captions exist puts them out of step.
- Burned-in captions are part of the picture and show everywhere. An .srt file is separate text that the viewer can switch on or off, and YouTube accepts it.
- White text with an outline or a box behind it stays readable over any footage.
What an iPhone can do on its own
Photos has a Caption field under each video, but it is a note for search. Nothing you type there appears on the picture.
Live Captions, in Accessibility settings, shows captions on screen for whatever is playing. It is there for the person watching on that phone. It does not write anything into the video file, so nothing travels with the video when you post it.
iMovie can put words on the picture. Tap a clip, tap Titles, pick a style, and type. Apple documents the steps in Add text to your iMovie project. That is the whole method: every line is typed by hand and timed by splitting the clip where the line should change. It works for a title card and a few labels. For two minutes of speech it is a few dozen splits and a lot of typing.
Appleās Clips app used to do this automatically with Live Titles. Apple stopped updating Clips and removed it from the App Store for new users in October 2025, so it is not an option on a phone that never had it.
How auto-captions work
An auto-caption tool does two jobs. It transcribes the speech into words, and it records when each word was spoken. The second half is what lets a caption appear on the word and leave on the pause.
Both jobs are done by a speech model, and it helps to know where one runs. Browser tools and many apps upload your video to a server first. Others download the model once and run it on the phone, so the footage stays where it is. Neither is more accurate by nature. The difference is where your video goes, and whether it works on a slow connection.
Every speech model gets some things wrong, and they are predictable: names, brand and product terms, numbers, and anything said while switching languages. Treat the first result as a draft.
How to add captions to a take
This is the workflow in Cutroom, an iPhone editor for talking videos. Captions are generated on the phone, and the caption model downloads the first time you use it.
- Finish the cuts first. Trim, split, and remove the pauses and filler sounds before you caption. If you edit afterwards, Cutroom tells you the captions are out of date and asks you to regenerate them. The draft you corrected is kept until the new one succeeds.
- Open Captions and pick the spoken language. English, Taglish, or Auto. Taglish is for Filipino, English, or the two mixed: English stays English and Tagalog stays Tagalog.
- Add names and terms to the glossary. It is optional, and it is the cheapest fix there is. A product name the model has never seen will be spelled wrong every time unless you tell it.
- Generate. The captions appear over the preview, and the phrase list, the look settings, and .srt export appear with them.
- Read every phrase once. Search the phrases for a word you know is risky, tap a phrase to fix its text, and change its start or end time if it arrives early or hangs on too long.
- Choose a look, then export. The captions are drawn into the exported video.
Auto Edit, above the timeline, runs the cuts, the voice cleanup, and the captions in one step. It gets you to step 5 faster. It does not replace step 5.
Choosing how the captions look
Start with the mode, because it changes how the video feels more than any colour does.
| Mode | What it shows | Suits |
|---|---|---|
| Dynamic | One to three centred words, highlighted as they are spoken | Reels, Shorts, TikTok, fast delivery |
| Classic | Longer phrases near the bottom of the frame | Lessons, interviews, landscape YouTube |
Four presets cover the usual cases: Clean, Minimal, Podcast, and Viral. From there you can set the style, the font, the text colour, the size, and what sits behind the words: nothing, an outline, a shadow, or a box. All caps and hiding punctuation are separate switches, and so is the highlight colour for the word being spoken.
Two things matter more than taste.
Contrast. Your background changes from frame to frame, and plain white text disappears over a white wall. An outline or a box keeps the words readable on any footage. Coloured text needs one of them even more than white does.
Position. Reels, Shorts, and TikTok cover the bottom of the frame with the description and the right edge with buttons. Keep captions clear of both. In Cutroom you drag the captions up or down on the preview, so you can see what they sit on top of.
Burned-in captions or an .srt file
Burned-in captions are pixels. They are part of the picture, they look the way you styled them, and they show on every platform with no setting to turn on. That is what you want for vertical short-form video.
An .srt file is plain text with times: each line says when a caption starts, when it ends, and what it reads. The player draws it, the viewer can turn it off, and the platform can read the words. YouTube lists SubRip (.srt) among its supported subtitle files, and desktop editors import it too.
You do not have to choose one. Export the video with captions burned in for the vertical cut, and export the .srt from the same captions for a long-form upload. In Cutroom, both are in the Captions sheet, next to Copy transcript, which gives you the text for a description or a post. Regenerate stale captions before exporting an .srt, so its times match the edit.
What to check before you post
- Names and numbers. Read them, do not skim them. A misspelled name is the mistake people notice.
- The first three seconds. The first caption should be on screen with the first word. A late first caption reads as a video with no captions.
- The bottom of the frame. Watch the export once on the platform, or leave room for its buttons before you export.
- Mixed-language lines. If you switched languages mid-sentence, check that the line kept both, in the words you used.
- One full watch with the sound off. If you can follow the video, so can a viewer on a train.
Questions people ask
Can the iPhone add captions to a video automatically?
Not with the apps that come on it. Photos and iMovie do not transcribe speech, and Live Captions only displays text during playback. You need an app that generates captions from the audio.
Do I have to upload my video to caption it?
No. Web tools need the upload because the transcription runs on their servers. Apps that run the model on the phone, Cutroom among them, never send the footage anywhere.
How accurate are auto-captions?
Accurate enough to save the typing, not accurate enough to post unread. Clear speech in one language into a decent microphone comes out mostly right. Names, jargon, background noise, and fast switching between languages come out worse. A glossary takes care of the names and terms; the rest is the read-through.
Should I use all caps?
For one to three words at a time, all caps reads fine and is the common style for short-form video. For full sentences, sentence case is easier to read. Match it to the mode: caps for Dynamic, sentence case for Classic.
Can I get the captions as text?
Yes. In Cutroom, Copy transcript puts the full text on the clipboard, and Export .srt saves a timed file that YouTube and most editors accept.
Where Cutroom fits
Cutroom is an iPhone editor for talking videos: it cuts the pauses and filler sounds, cleans up the voice, and captions the result, all on the phone. Editing and previewing are free, and the first export is free with a Made with Cutroom watermark. It is available on the App Store for iPhone.