Generate Auto-Captions in CapCut
Captions are the single highest-return edit you can make. Most people scroll social feeds with the sound off, and even with sound on, captions boost how long viewers stay. The old way meant typing every word and timing it by hand. CapCut's auto-captions do a full video in under a minute. This lesson gets you from raw footage to styled, readable captions.
What You'll Learn
- How to generate auto-captions in CapCut in one click
- How to fix transcription errors quickly
- How to style captions so they are readable and on-brand
- How to use the popular word-by-word "highlight" caption look
Generate the Captions
With your clip on the timeline:
- Open the Captions (or Text) menu, then choose Auto-captions.
- Select the spoken language. Pick the language that is actually spoken in the clip, not your interface language.
- Click Generate.
In seconds, CapCut transcribes the whole video and lays the text onto the timeline as a caption track, already timed to your speech. This is speech-to-text AI at work. It listens to the audio waveform and matches it to words.
That first pass will be roughly 90 to 95% accurate for clear speech in a common language. The remaining few percent is your job to fix.
Fix the Errors Fast
Auto-captions get tripped up by predictable things: proper names, brand names, jargon, homophones (their vs there), and unclear audio. Do not read the whole transcript word by word. Instead:
- Open the caption text list (CapCut shows all caption segments as an editable list).
- Play the video once at normal speed and watch the captions. Your eye catches errors faster than reading cold.
- Click any wrong segment and retype it.
For names and terms you know will appear, fix them once and use CapCut's find-and-replace if available, or just scan for that word. If you have a lot of specialized terms, the next lesson shows how AI can clean the whole transcript in one shot.
A quick accuracy tip: better audio means better captions. If your recording has background noise, the transcription suffers. You will clean audio in Module 4, but know that clean input is the real fix.
Style Captions So People Actually Read Them
Default captions are often too small, low-contrast, or badly placed. Readable captions follow a few rules.
Size and contrast. Text should be large enough to read on a phone held at arm's length. Use a solid color with a contrasting outline or background box. White text with a thin black outline is the safe, universal choice. Never put plain white text on a bright or busy background with no outline.
Position. Keep captions in the lower third but not so low that platform UI (the TikTok caption bar, YouTube progress bar) covers them. On 9:16 vertical video, leave room at the very bottom.
Font. Pick one clean, bold sans-serif and stick with it. A bold font reads better at a glance than a thin or decorative one.
Length on screen. One or two short lines at a time. Long paragraphs of caption are unreadable while video moves.
To apply a style: select a caption, open the style panel, set the font, size, color, and outline, then use Apply to all so every segment matches. Do this once and the whole video is consistent.
The Word-by-Word Highlight Look
You have seen it: captions where each word pops or changes color as it is spoken. This "karaoke" or highlight style is popular because it pulls the eye along.
In CapCut, look for Auto-captions styles or the Highlight / Karaoke caption templates. Many are one click. Apply one and CapCut animates the words in time with the audio automatically.
Use this style for fast, energetic short-form content. For a calm explainer or a formal presentation, plain clean captions read better. Match the style to the mood.
Let AI Help You Decide
Not sure which caption style fits your video? Ask. Paste this into ChatGPT, Claude, or Gemini:
I made a [topic] video for [platform, e.g. TikTok / YouTube / a class presentation]. The tone is [energetic / calm / professional]. Recommend a caption style: font weight, one or two lines, color, and whether to use animated word-by-word highlighting. Explain why in one sentence each.
You will get a quick, reasoned recommendation instead of guessing.
Accessibility Matters
Captions are not just for silent scrolling. They make your video usable for deaf and hard-of-hearing viewers, and for anyone in a noisy or quiet room. Accurate captions are an accessibility feature, which is one more reason to proofread rather than ship the raw transcription. Many schools and workplaces expect captioned video for this reason.
Try It Now
- Generate auto-captions on your clip.
- Play it once and fix every error you spot, especially names and terms.
- Style the captions: bold sans-serif, white with a black outline, lower third. Apply to all.
- Try one animated highlight style and compare it to the plain style. Keep whichever fits your video.
Key Takeaways
- Auto-captions transcribe your whole video in under a minute at 90 to 95% accuracy.
- Always proofread; names, jargon, and homophones are the usual errors.
- Readable captions are bold, high-contrast, one or two short lines, in the safe zone.
- Word-by-word highlight styles suit energetic short-form; plain captions suit calm or formal video.
- Captions are an accessibility feature, not just an engagement trick.