tonboxai.online/blog/how-to-add-captions-to-ai-voiceover-videos-in-capcut

How to Add Captions to AI Voiceover Videos in CapCut

Published 24 October 2026 · Written by the TonBox AI team

Captions help viewers follow your video in noisy places, on mute, or when they are learning a language. They also keep attention, because people read along while they listen. If you make videos with an AI voiceover, adding captions is quick, and most editors, including CapCut, have a built-in way to do it. The exact menu names may change between versions, so this guide focuses on the steps rather than button labels.

1. Prepare your audio first

Generate the voiceover in the voice studio and download the MP3. Listen to the full file once before you start editing. Fixing a mispronounced word in the script is much easier than fixing it after you have built captions and visuals around it.

2. Start a new project and import the audio

Create a new project, set the aspect ratio for where you will publish (vertical 9:16 for Shorts and Reels, horizontal 16:9 for regular YouTube videos), and import your MP3. Place the audio on the timeline first. Everything else, including visuals and captions, is built around it.

3. Add your visuals

Add images or clips and cut to a new visual whenever the voice moves to a new idea, usually every two to four seconds. Make sure you have the right to use every image and clip. Using your own screenshots, stock footage with a clear licence, or material you filmed yourself keeps your channel safe.

4. Generate captions automatically

Look for an automatic captions or speech-to-text option in the text or captions menu. Choose the language of your audio and run it. The editor listens to your voiceover and creates timed text. Because AI voices are clear and steady, automatic captions are usually quite accurate, though not perfect.

If you wrote the script yourself, some editors also let you paste the script and align it to the audio. This can be more accurate than automatic recognition, especially for names and unusual words.

5. Correct the captions

Read through every caption while the video plays. Check names, numbers, brand names and words in other languages, which are the most common mistakes. If your script mixes Hindi and English words, look carefully at those lines. Fix the text, and adjust timing if a caption appears too early or stays too long.

6. Keep lines short

On a phone screen, two short lines are easier to read than one long block. Aim for a few words per caption, split at natural phrase breaks. Avoid breaking a name or a number across two captions.

7. Style captions for readability

  • Choose a clear, bold font and a size that is readable on a small screen.
  • Use strong contrast, such as white text with a dark outline or a soft background box.
  • Keep captions away from the very top and bottom of the frame, where the platform's own buttons and text can cover them.
  • Use one style for the whole video so it looks consistent.
  • Highlight a keyword occasionally for emphasis, but do not overdo it.

8. Check the whole video on a phone

Export a test version and watch it on an actual phone. Look for captions that are cut off, too small, hidden behind interface elements, or out of sync with the voice. Fix these before the final export.

9. Export and upload

Export at a resolution that suits your platform, and check the file one more time before uploading. If the platform lets you upload a separate caption file, you can also do that, which helps with accessibility and search.

Quick checklist

  • Voiceover listened to before editing.
  • Right aspect ratio chosen at the start.
  • Captions generated, then corrected line by line.
  • Short lines, readable font, strong contrast.
  • Test export checked on a phone.

Good captions take only a few extra minutes and make your video easier to watch for everyone. For more on settings that make the voice itself sound better, read our guide to the best text to speech settings for YouTube Shorts.

← All guides