How do I add an AI voiceover to a CapCut video?
The fastest way is to generate the voice outside CapCut and import it as an audio file. Write your script, turn it into speech with Speechma’s free text to speech tool, download the MP3, then import that MP3 into your CapCut project and line it up with your clips. You get access to 1,700+ voices in 75+ languages and locales, and the file is yours to keep and reuse in any editor.
Why not just record your own voice? Plenty of creators do, but a home recording often picks up fridge hum, traffic and room echo. An AI voiceover sounds the same every time, so a 10-part YouTube Shorts series or a Shopify product reel keeps a consistent narrator from the first clip to the last. If you change a line later, you regenerate one sentence instead of re-recording the whole take.
Step-by-step: add a Speechma voiceover to CapCut
- Write a script that fits your clipPlan about 150 words per minute of video at normal speed. A 30-second TikTok or Reel needs roughly 70–80 words. Write numbers the way they should be spoken (“twenty dollars,” not “$20”).
- Paste it into Speechma and pick a voiceOpen Speechma text to speech, paste up to 2,000 characters, and filter the voice list by language and country. Click a few previews; an upbeat US voice suits product reels, a calm one suits tutorials.
- Fine-tune speed, pitch and pausesOpen Voice Effects to nudge speed, pitch and volume. Use the Short, Medium or Long pause buttons to insert a real pause such as
[break 1s], and let punctuation shape the rest: a comma gives a short breath and a full stop a slightly longer one. - Generate and download the MP3Type the 5-digit security code, click Generate Audio, listen back and download the MP3. If you made several parts, use “download all” to grab them together as a ZIP.
- Import the MP3 into CapCutOn CapCut desktop or web, open your project, click Import in the media panel and choose the MP3. On the CapCut mobile app, tap Audio in the bottom toolbar and use the option to add audio from your device (often under Sounds or an Import/“Your sounds” section, depending on your app version).
- Drag it onto the timeline and syncDrop the audio on a track below your video. Use Split to cut the voiceover at sentence breaks, then slide each piece so it lands on the right shot. Trim the ends by dragging the edges of the audio clip.
- Balance the volume and exportLower background music so the voice sits on top, or use CapCut’s loudness normalization option to even things out. Preview the whole video with headphones, then click Export.
Which voice settings work best for CapCut videos?
These are starting points, not rules. Generate a test line, listen inside CapCut with your music underneath, and adjust.
| Video type | Voice style | Speed | Pitch | Pacing tip |
|---|---|---|---|---|
| TikTok / Reels hook | Energetic US English | Slightly faster | Default or a touch higher | Short sentences, exclamation marks for punch |
| Product demo (Shopify, Amazon) | Friendly, clear | Default | Default | Commas between features so viewers can read on-screen text |
| Tutorial / how-to | Calm, neutral | Default or slightly slower | Default | Semicolons between steps for a medium pause |
| Storytime / Reddit-style | Warm narrator | Default | Slightly lower | Vary sentence length to avoid monotone |
| Real estate walkthrough | Confident, polished | Slightly slower | Slightly lower | One room per sentence, matched to each clip |
CapCut accepts common audio formats including MP3, which is what Speechma downloads.
Troubleshooting: voiceover problems in CapCut
- The voice is out of sync with the visuals. Split the audio at each sentence and move the pieces individually. It is easier than stretching clips to fit one long take.
- The voiceover is too long for the clip. Cut words rather than speeding the voice up a lot; fast AI speech sounds rushed. Small speed increases in Speechma are fine.
- The music drowns out the voice. Select the music track and lower its volume, or add fade in/out so the voice stands out at key moments.
- I can’t find the MP3 on my phone. Mobile browsers usually save downloads to the Files app (iPhone) or the Downloads folder (Android). Move it somewhere easy to find before you open CapCut.
- A word is mispronounced. Spell it phonetically in your script (for example “Speech-ma”), regenerate just that sentence, and replace the piece on the timeline.
- My script is over 2,000 characters. Split it into parts at paragraph breaks and generate each one, or use SpeechmaPRO.com for 10,000 characters at once with no ads or captcha.
Pro tip: generate one sentence per clip
For fast-cut edits, generate each sentence as its own short MP3 and name the files 01, 02, 03 before importing. Each file maps to one shot, so you rarely need to split anything. When the client asks to change line four, you regenerate only that file. For narrated videos built on our voices inside your own app or workflow, the Speechma API starts at $9/month.
Frequently asked questions
Is it free to add an AI voiceover to CapCut with Speechma?
Can I use the voiceover in monetized TikTok or YouTube videos?
What audio format should I import into CapCut?
Why use Speechma instead of CapCut’s built-in text to speech?
How do I make the AI voiceover sound less robotic in CapCut?
Does this work on the CapCut phone app?
Make better voiceovers, faster
- New voices and features the day they launch
- Script templates and tricks that make AI voices sound human
- Creator stories, community picks and early access
