CapCut workflow

Creator notes

Step-by-step CapCut workflow: export AI voiceover MP3 + SRT, drop them on the timeline, sync captions, and export Shorts-ready video.

Workflow9 min readAug 3, 2026

← All articles

How to add AI voiceover and SRT captions in CapCut

Most Shorts get finished in CapCut. The voiceover should land there as a clean MP3, and captions should come from an SRT you already trust — not from a second transcription pass if you can avoid it.

This guide is the CapCut packaging loop only: generate audio + captions elsewhere, then assemble the edit. It is not another “best AI voice” roundup.

What you need before CapCut

  1. A Shorts-length script (speakable lines, not an essay)
  2. An MP3 voiceover
  3. An SRT file timed to that same take
  4. B-roll, screen recordings, or stills for the visuals

In Shorts Voice Studio you can generate the MP3 and optional SRT together. Keep each generate under 700 characters; split longer ideas into sections and stitch in CapCut.

CapCut steps: audio first

  1. Create a new project (9:16 for Shorts / Reels / TikTok)
  2. Import the MP3 and drop it on the timeline
  3. Build visuals to the voice — cut on breaths and idea breaks
  4. Keep music quieter than the narration if you add a track

CapCut steps: import the SRT

Exact menus move between CapCut desktop and mobile, but the job is the same: import captions from file, then style them for a phone screen.

  1. Import or add captions from your SRT file
  2. Check the first hook line lands in the opening two seconds
  3. Fix names, numbers, and brand spellings once
  4. Move captions away from faces and UI you need people to see
  5. Export, then watch once on a phone before you upload

Burn-in vs upload later

  • Burn captions into the CapCut export for TikTok and Reels-style feeds
  • On YouTube, you can burn in or upload the same SRT in YouTube Studio
  • One SRT source keeps both paths honest — do not retype captions from scratch

Common CapCut mistakes

  • Building the whole video before locking the voice take
  • Letting music drown the offer or tip line
  • Leaving auto-captions when you already have a good SRT
  • Regenerating voiceovers mid-edit because the script was never spoken aloud once
Lock the voice track first. CapCut gets faster when the audio is the boss.

Put this into practice in Shorts Voice Studio.

Generate MP3 + SRT for CapCut
CapCut AI voiceoverCapCut SRT captionsAI voiceover CapCutYouTube Shorts CapCut