How to add subtitles in CapCut
CapCut transcribes a clip in seconds. The five minutes after that — corrections, line breaks, styling and placement — are what separate readable subtitles from a wall of text.
7 minute read
What CapCut does well here, and what it leaves to you
CapCut's auto-captions are quick and, on clear audio, accurate enough that most of your work is checking rather than typing. The app transcribes the spoken audio, lays timed caption events onto the timeline, lets you restyle every line at once, and burns them into the exported file. What it does not do is judge. It will not notice that it heard your company name wrong, that a line runs to three cramped rows, or that the caption block is sitting exactly where TikTok draws its own buttons. Everything below is written for a vertical clip destined for TikTok, Reels or Shorts, and the order matters: get the words right, then the timing, then the look, then the placement. Styling first is the most common way to end up redoing the job.
Step one: import the clip and generate the captions
Bring your video in, then find the captions entry point. On mobile it is a Captions button in the bottom editing toolbar, and you may have to scroll the toolbar along to reach it. On desktop and web the same feature lives under the Text tab as Auto captions. CapCut's own help centre also describes the action as recognising subtitles, so expect the wording to shift a little between versions and platforms rather than assuming a screenshot from last year still matches. Whichever route you take, the important step is the one people skip: choose the spoken language, and the regional variant if you are offered one, before generating. Transcription runs in the cloud, so you need a connection, and it will take seconds rather than minutes on a short clip.
Step two: fix the words that carry weight
Speech recognition is reliably good at ordinary sentences and reliably poor at exactly the words you care about most: names, brands, product terms, acronyms and figures. CapCut publishes a help article specifically on fixing inaccurate auto-captions, which tells you how routine this is. Read the transcript against the audio once, and prioritise the first line, since a misspelled name in the opening second undermines everything after it. CapCut also supports batch editing for recognised captions, covered in its own help centre, which lets you change text across the generated lines rather than opening each one, and that is the fastest way to correct a term you say fifteen times. Note the limit: batch editing applies to auto-generated captions, not to text layers you typed in by hand.
Step three: break the lines so they can actually be read
Auto-captions arrive broken where the transcriber found pauses, which is not the same as where a reader needs breaks. Subtitling has published numbers worth borrowing. Netflix's English style guide allows a maximum of two lines, up to 42 characters per line, and a reading speed of up to 20 characters per second for adult content, with a minimum of five-sixths of a second and a maximum of seven seconds on screen for any one subtitle event. Twenty characters a second is roughly three words a second, and that is a ceiling, not a target. On a phone-sized vertical clip, one or two short lines per card reads far better than a dense block. Split at natural clause boundaries, never leave a single word stranded on its own line, and make sure each card lands with the words as they are spoken.
Step four: style for legibility first, personality second
Select the caption track to restyle every line together rather than one at a time. The choices that matter are unglamorous: a heavy sans-serif rather than a thin or script face, a size that looks slightly too large on your monitor, and either a stroke or a semi-opaque background box so the text stays readable when the footage behind it changes. WCAG 2.2 asks for a contrast ratio of 4.5:1, relaxing to 3:1 for large text at 18 point or 14 point bold, and caption type on a phone usually counts as large. CapCut's animated caption templates, including the word-by-word highlight styles, genuinely help hold attention on short-form. Just pick one and keep it, because a different animation on every clip reads as noise rather than as a channel.
Step five: move the captions out of the interface
CapCut previews your clip on a clean 9:16 canvas. TikTok, Reels and Shorts do not. Each of them draws a caption, an audio label, a profile control and a column of buttons over your frame, and the default caption position in an editor often sits underneath them. Drag the caption block up into the middle third, keep it clear of the speaker's mouth, and leave the bottom of the frame empty. If you are posting the same clip to more than one app, check it against each layout rather than trusting one preview, because the covered area differs between them and shifts with app updates. This is a thirty-second check that prevents the single most common short-form mistake, which is subtitles that are perfectly styled and completely hidden.
Step six: export burned in, or as a subtitle file
For social platforms, export the video with the captions baked into the picture. Burned-in text renders identically everywhere, cannot be toggled off by the viewer, and does not depend on each app's caption system behaving. CapCut also supports exporting the subtitles as a separate file, which is worth doing when the same footage is going to a long-form YouTube upload where a real caption track feeds search and translation and lets viewers restyle it. One honest caveat on the commercial side: some caption and export features sit behind CapCut Pro, and what the free tier includes has changed more than once. Rather than trusting any number you read online, this article included, look at what the app tells you when you press Generate or Export.
When this workflow stops being the right one
CapCut is a strong manual editor, and the process above takes a few minutes per clip. That is fine for one clip and it is fine for five. The maths changes when the input is a two-hour podcast and the output is twenty verticals a week, because you are then repeating the same corrections, the same styling and the same repositioning twenty times over. At that point the bottleneck is not the caption tool, it is the per-clip repetition. That is the job a clipping tool like FrameOS is built around: one long recording in, a batch of reframed and captioned verticals out, with the styling decided once for the whole batch. Different tool for a different scale of problem, and worth knowing about before you spend a Sunday on it.
FAQ
How do you add automatic subtitles in CapCut?
Import the clip, open Captions from the bottom toolbar on mobile or Auto captions under the Text tab on desktop and web, choose the spoken language, and generate. CapCut transcribes the audio and places timed caption lines on the timeline. Its help centre also refers to this as recognising subtitles, so the exact label varies between versions.
Can you edit CapCut auto-captions after generating them?
Yes, and you should. CapCut publishes help articles both on fixing inaccurate auto-captions and on batch editing recognised captions, which lets you correct a repeated word across the whole track instead of line by line. Prioritise names, brands, product terms and numbers, since those are where speech recognition fails most often.
How do I stop CapCut subtitles being covered on TikTok?
Move them up. CapCut previews a clean 9:16 frame, but TikTok, Reels and Shorts each overlay captions, buttons and an audio label on top of your video, mostly along the bottom. Place the caption block in the middle third, keep it off the speaker's mouth, and check the clip against each app's layout before posting.
Can CapCut export subtitles as an SRT file?
CapCut supports exporting subtitles as a separate file as well as burning them into the video, though some caption and export features sit behind CapCut Pro and the free tier's limits have changed over time. For social clips, burned-in is the right default anyway; a subtitle file is more useful for a long-form YouTube upload.
Sources
- How Do I Fix Inaccurate Auto-Captions in CapCut? · CapCut Help Centre
- How Should I Batch Edit to Recognize Subtitle/Lyrics Text in CapCut? · CapCut Help Centre
- How do I recognise subtitles? · CapCut Help Centre
- English (USA) Timed Text Style Guide · Netflix Partner Help Center
- Web Content Accessibility Guidelines (WCAG) 2.2 · W3C
Related pages
Keep reading
More Captions & Subtitles →How to translate video captions
Auto-translated captions inherit every error in the auto-transcript underneath them, and translated text runs longer than the English it replaced. Here is the order of operations that avoids both.
8 minute readWhat are burned-in captions?
Burned-in captions are subtitles baked into the video itself, so they always appear and can't be turned off. Here's when to use them versus closed captions.
5 minute readSubtitles vs captions: what's the difference?
Subtitles and captions are often used interchangeably, but they're not the same. Here's the actual difference — and which one you need for short-form video.
5 minute readTurn one long video into a week of posts
300 credits for 3 days · no card.
Start free