Before you start
CapCut on desktop and CapCut on a phone are close cousins, not twins: the tools have the same names but live in different places, and tracking is better on the phone. The steps below are written for the desktop app with the phone equivalents noted. One thing to know about CapCut specifically: an effect applies to the whole frame unless you mask it, so covering a mouth means a masked duplicate or a sticker, not a blur on the main clip.
Plan the order: captions to find the words, then the audio, then the cover, then export, and let the platform caption the clean file afterwards. Features move between versions and some sounds and effects are limited to CapCut Pro, so if a menu is not where this says, search for it in the app's search box.
Find the swears with auto captions
- Desktop: Text, Auto captions, Generate. Phone: Text, Auto captions.
- Read down the caption track. Clicking a caption jumps the playhead to it; note the time of every swear, including plurals and -ing forms.
- Before you export, edit or delete the caption track, or the word is burned into the picture even after you have muted it.
Auto captions miss shouted and overlapping words. Scrub the waveform around each one you find, and listen to the loud moments twice.
00:42no way he just did that
03:17okay that was brilliant
07:55chat, I am not doing that again
12:08what the was that
Bleep a word
- Right-click the clip and choose Separate audio, so your cuts only touch the sound.
- Put the playhead at the start of the word and press Ctrl+B (Cmd+B on Mac) to split. Do the same at the end.
- Select the slice and set its Volume to the minimum in the panel on the right (or right-click, Mute).
- Open Audio, Sound effects and search "beep" or "censor". Drag one under the gap, trim it to the slice and lower its volume to sit level with the voice.
- Add a short fade in and out on the slices either side (Audio, Basic, Fade) to avoid clicks.
- Phone: tap the clip, Split; Volume to 0; Audio, Sounds, search "censor beep".
Mute instead
Skip the sound effect. Same splits, volume to zero, short fades, and move on. The whole mix drops for the word, music included; if the music is on its own track, mute only the voice track's slice and the bed carries through.
Cover the mouth
Blur or mosaic
- Copy the video clip and paste it on the track above, then split the copy at the word's start and end and delete the rest, so the duplicate exists only for the word.
- With the duplicate selected, open Effects, Video effects, search Mosaic or Blur, and apply it. The whole duplicate is now blurred.
- Open Mask on the duplicate and choose the circle. Size and position it over the mouth and feather the edge. Now only the mouth shows the blurred layer.
- Add a keyframe on the mask position at the start of the word and another at the end, with more in between if the head moves. If your version has Tracking (the phone app has it for stickers and some effects), use that instead of keyframes.
- Check at full size; the cover must hold for every frame of the word.
Sticker
Open Stickers, pick an emoji, drop it over the mouth, trim it to the word and keyframe or track it. It is faster and reads as a joke, which suits some clips and not others.
Check it before you post
Play the clip back at full size with the sound on, then once more on mute. The first pass catches a bleep that starts late or a slice that clips the word before it; the second shows you whether the mouth still gives the word away, which on a close-up it will unless it is covered. Look at the caption track too: if you generated captions before the edit, the swear is still written there. Export at the clip's own resolution and frame rate (CapCut's defaults are fine), and let TikTok or Instagram caption the clean file; their captions are made from the audio they receive, so a word that is not in the audio is not in the text.
The faster way
For one clip with one swear, the steps above take a few minutes. For a batch of clips, or a long video, export from CapCut as you normally would, upload the file to Redactify, pick the preset for where it is going (TikTok Safe, YouTube Ad-Friendly, Instagram Safe), turn on mouth blur if anyone is on camera, mark each face once, check the list and download. Then post the clean file and let the platform caption it; the captions come from the clean audio, so the word does not come back as text. Bulk upload on Pro and Premium queues a whole batch at once. If you still want captions styled in CapCut, bring the clean file back in and caption that.
Audio costs 1 credit a minute and video with mouth blur 2; new accounts get 30 credits free.