How to add subtitles in CapCut (and when not to)

CapCut can auto-caption a clip in a couple of taps. The catch is what's behind that feature today, and how much you can actually edit once it's there. Here are three ways to get captions onto a CapCut timeline, in order of how much of your evening they cost you.

Option 1: CapCut's auto captions

Add auto captions from the Text menu and CapCut transcribes the clip on-device. It's fast and it's built in, which is exactly why it's the default choice for most people. Two things bite in practice: accuracy drops on background music or overlapping speakers, and full editing control — no watermark, higher export resolution, some styling options — moved behind CapCut's paid tier during the 2026 pricing restructure that split the old single Pro plan into a cheaper Standard tier and a pricier new Pro tier.

Option 2: type captions manually

Full control, zero dependency on any subscription — and a ten-minute video is a genuine hour or more of scrubbing and retyping. Fine for a 15-second clip. Not a workflow you want to repeat every week if you're publishing on any kind of schedule.

Option 3: transcribe elsewhere, import the SRT

CapCut imports SRT files directly, and nothing about that import is behind a paywall. Generate the transcript with word-level timing somewhere free, edit the lines and line breaks there, export SRT, and drop it onto your CapCut timeline. You keep CapCut for the actual edit — trims, transitions, effects — and get subtitles that were never gated by which tier you're on.

1. Extract your audio or drop in the raw video file (audio is decoded in your browser, so nothing large has to upload).
2. Wait for transcription and forced alignment — every word gets its own timestamp, not just each line.
3. Set words per line, or switch to natural sentence breaks, and split or join any line that reads awkwardly.
4. Export SRT.
5. In CapCut: tap the clip, then Text → Add subtitle → Import, and select the file.

Common problems

• Auto-captions miss words on music-heavy clips: a forced-alignment pass against a clean transcript is more accurate than recognition running on the fly.
• Captions look right in the editor but the export is watermarked or capped at 1080p: that's the free-tier export limit, not a subtitle problem — importing SRT doesn't get around it, since that's separate from captioning.
• Line breaks land mid-phrase: CapCut's auto-segmentation breaks on a fixed word count. Natural-sentence segmentation elsewhere reads noticeably better.

Frequently asked questions

Does CapCut's auto-caption feature cost money?

The basic auto-caption pass is free to generate, but full export quality and some editing controls sit behind CapCut's paid tiers. Importing an SRT you generated elsewhere is not gated the same way.

Can I edit CapCut's auto-generated captions afterward?

Yes, but line-by-line, inside CapCut's caption editor, with no per-word timestamps to fall back on — so splitting a line means guessing where the new one should start.

Will an imported SRT keep its formatting in CapCut?

SRT carries no styling by design — just text and timecodes. Apply CapCut's caption style to the imported track after import; that part isn't affected by the file's own (lack of) formatting.