CapCut’s auto captions turn spoken audio on your timeline into timed subtitle blocks you can edit, style, and export for social video—TikTok, YouTube, Reels, Shorts, and similar feeds.
Editors with an existing cut use this path when the job is generate → correct → export captions for platform posts, not a full timeline rebuild. You will get the Auto Captions steps early, then burn-in vs editable trade-offs for how you ship, plus timing and style checks, soft free-to-start access notes, and how this captions layer differs from a generic generator primer. For that primer, see CapCut’s video caption generator; see CapCut’s AI video editor guide for the full timeline-plus-assists overview.
- What CapCut auto captions do for social cuts
- Step-by-step: auto-generate, edit, and export captions
- Burn-in vs editable captions: which output to ship
- Free vs access: what to check after the first caption pass
- Timing, style, and readability checks that protect completion
- Output specs: burn-in export, sidecar files, and font/lyric licensing
- Common failure modes and how to fix them
- Conclusion
- FAQ
What CapCut auto captions do for social cuts
CapCut works as an auto captions ai video editor when footage already sits on a timeline and the job is readable subtitles for TikTok, YouTube, Reels, Shorts, or other social posts—not a from-scratch edit rebuild. Auto Captions / Recognise Subtitles turns speech into timed text blocks you still own: fix names, retime lines, apply a TikTok-friendly caption style, then export.
Use this path when speech clarity on mute-first feeds (or captioned long-form) is the bottleneck: generate the draft, edit wording and timing, style for phone glance reading, and export. Burn-in vs editable / sidecar is the export choice later—run the steps first so you have clean text before you lock how it ships.
Live CapCut help (early 2026) lists Auto Caption / Recognise Subtitles on Web, Desktop, and Mobile. Advanced extras—speaker differentiation, punctuation polish, or batch .srt export—may appear on Desktop or Web by region and version. Treat the Captions panel in your build as live. CapCut’s auto video editor is a broader auto-edit conversion landing; stay here when captions on an existing cut are the decision.
Step-by-step: auto-generate, edit, and export captions
Labels can shift by version and region. If a menu name differs, follow Captions / Auto Captions / Recognise and Export in your build. For a dedicated captions product surface, CapCut’s AI caption generator maps the same generate → edit → export loop on Web.
Shared path (all surfaces)
- 1
- Open the project and ensure the spoken track is audible (mute competing music if needed). 2
- Open Captions → Auto Captions → Recognise (or Generate). 3
- Choose the spoken language → Generate. 4
- Edit text, split long blocks, retime ends, and apply style in the caption panel / timeline. 5
- Preview on a phone-sized frame, then Export—at export you choose burn-in into the picture or keep editable / sidecar text when your build lists it (details in the next section).
Platform differences (Web / Desktop / Mobile)
Publish checklist before you leave Export: language matched speech; names and numbers corrected; line length readable on phone; then pick burn-in vs editable on purpose and apply the Output branding check.
Burn-in vs editable captions: which output to ship
Burn-in captions bake words into the picture so mute-first viewers always see them; editable or sidecar captions stay as text you can fix without re-encoding. Choose burn-in for TikTok, Reels, and Shorts when guaranteed on-screen speech matters; choose sidecar when YouTube or reviewers need toggleable CC and cheaper wording changes.
Burn-in (open) captions render text into the picture at export. Viewers cannot toggle them off; any wording change needs a re-edit and re-export. Editable / sidecar captions stay as project text (and, where your surface supports it, as .srt / .txt or platform closed captions). Styling you applied in CapCut usually does not travel with a basic SRT—fonts, animations, and positions stay with burn-in or the CapCut project.
Many teams ship a burn-in social master for TikTok / Reels / Shorts plus a sidecar long-form master for YouTube—choose before Export so you do not stack open captions on top of platform CC by accident.
Free vs access: what to check after the first caption pass
CapCut offers a free-to-start path for auto captions on many surfaces—for a first generate → correct → export loop, not an unlimited forever promise.
Keep the first test under ~60 seconds of clear speech so one try teaches generate → correct → choose burn-in or sidecar, then check live unlocks.
Timing, style, and readability checks that protect completion
Readable captions protect muted-feed completion more than fancy animation: keep lines short, clear of platform chrome, high-contrast on busy B-roll, and spell names and numbers correctly before burn-in. Run the checks below before you lock open captions into the picture.
When a retime is mandatory. After you shorten, reorder, or punch-in the picture, caption blocks often drift. If only timing slipped, drag ends on the timeline; if many lines are wrong, delete captions, mute music, regenerate, then re-apply style. Do not burn-in a drifted pass.
Output specs: burn-in export, sidecar files, and font/lyric licensing
Burn-in export. Export the video with captions rendered into the frame when the destination is mute-first social or when the platform will not carry your CapCut styling via CC. Confirm resolution, aspect (often 9:16 for short-form), frame rate, and format in Advanced Settings when shown. Preview the burn-in on a phone before you ship.
Sidecar / editable export. Where Desktop or Web lists .srt / .txt (or platform CC upload), export the text track for YouTube, review rounds, or multi-language packs. Remember: basic SRT typically does not keep CapCut font, animation, or position—restyle on the destination or keep a CapCut project master.
Font, lyric, and commercial notes. Treat caption fonts, lyric/synced text overlays, and music under the spoken track as account- and license-dependent. CapCut tool pages may market broad commercial use; for Resource guidance, review the fonts and audio you actually used, stay on assets your brief allows, and do not promise unconditional commercial clearance from auto captions alone. Soft branding check: inspect the exported file rather than assuming tool-page “no watermarks” language always applies.
Common failure modes and how to fix them
When the cut is close, fix text and timing on the timeline. For the full edit overview, open CapCut’s AI video editor guide instead of stretching this captions Resource.
Conclusion
Generate Auto Captions on CapCut’s timeline, correct names and timing, and style for phone glance reading—then decide how the words ship.
Choose burn-in when mute-first TikTok, Reels, or Shorts feeds must always show the words; choose editable / sidecar when YouTube or reviewers need toggleable text and cheaper wording changes. Export the path that matches the platform before you scale the same style across a batch. For the broader timeline-plus-assists overview, see CapCut’s AI video editor guide (linked in Preface and FAQ 6).
FAQ
1. How accurate are CapCut auto captions, and what still needs a human pass?
CapCut auto captions are strong draft speech-to-text on clear single-speaker audio, but still need a human pass for names, numbers, brands, and noisy or overlapping speech. Accuracy varies with mic quality, accents, music under dialogue, and the language you select. Preview with sound and edit lines before burn-in—auto captions draft; they do not replace review.
2. Burn-in vs sidecar captions — which should I export for TikTok, Reels, or YouTube?
For TikTok, Reels, and many Shorts cuts, burn-in captions usually fit mute-first feeds where viewers cannot rely on toggle CC. For YouTube or review handoffs, editable / sidecar captions (SRT when listed) usually fit better so text can be fixed without re-encoding the picture. Pick one primary path per master to avoid stacking open captions on top of platform CC.
3. Which languages does CapCut auto captions support?
CapCut auto captions support multiple spoken languages on Web, Desktop, and Mobile—often marketed as 20+ on caption-tool pages when listed—but the set is region- and version-variable. Choose the spoken language before Generate, then spot-check a short sample. Do not assume every dialect or specialty term lands without edits.
4. Is CapCut auto captions free to generate and export?
CapCut offers a free-to-start auto captions path on many surfaces for generate, edit, and common exports when listed. Free is not unlimited: after the first pass, higher export settings, Pro-marked styles, or other gates may appear—apply the Output branding check. Try the loop first, then confirm live unlocks for client work.
5. When do I need to retime captions after editing the picture?
Retime captions whenever you shorten, reorder, split, or punch-in the picture after the caption pass—otherwise speech and text drift. If only a few lines slipped, drag block ends on the timeline; if many lines are wrong, delete captions, mute competing audio, and regenerate. Never burn-in a drifted pass for a platform cut you care about.
6. When should I use this captions workflow instead of CapCut’s full AI video editor guide?
Use this captions workflow when footage already sits on a timeline and the bottleneck is generating, editing, and exporting subtitles for social video—plus burn-in vs editable output when you ship. Open CapCut’s AI video editor guide for the broader edit pillar—import, multi-track timeline, and AI assists beyond captions. Captions depth lives here; full assists overview lives in that guide.