CapCut Video Translator: Subtitles, Dubbing & Fixes

To translate a video in CapCut, first decide what should change: the subtitles or the spoken voice. For translated text, open Captions → Auto captions and check the translation options. For dubbed speech, select a video clip on Desktop and look for Video translator in its right-hand Audio settings. That dubbing feature is region-dependent and may not appear in your account.

Review a short sample before translating the full project. A fluent-sounding voice can still say the wrong name, price or instruction.

Text and official sources checked September 5, 2026. The illustrations are editorial diagrams, not current screenshots of an account or proof of a successful translation. Follow the controls and availability messages in your version.

Choose subtitles or AI dubbing

What CapCut Video Translator does

CapCut's Video Translator page describes translated subtitles. Its separate AI Video Translator page describes translated speech, voice-style preservation and lip-sync. Similar names, different outputs.

Recognition errors can carry into either result. If a person's name is transcribed incorrectly, compare the translation with the recording and correct it. But don't assume editing a caption changes what a separate dubbing tool will say: that tool may recognize the audio independently.

Voice preservation and lip-sync are advertised capabilities, not guarantees. Listen for changes between clips and inspect the actual face movement. A natural voice alone does not prove the meaning is right.

Before you translate a real project

• Keep a separate project copy and the source media. Saving the current edit is not an independent backup; an exported video is not an editable-project copy.
• Make a short test project. Include an ordinary sentence, a difficult name, a number and a visible speaker if lip-sync matters. This bounds the test without assuming that selecting one clip limits every AI action to that clip.
• Check the speech source. Play it with headphones. Use the actual audio-source selector when offered. Playback mute alone does not reliably tell you which track recognition will process.
• Write down important terms. Keep names, brands, dates, units and approved phrases beside you during review. This is a review checklist, not an assumption that the translator accepts a terminology prompt.
• Confirm permission before uploading. Check that you may process the footage and voice, especially for client work. CapCut's materials and commercial-use guidance does not clear unrelated footage or somebody else's voice.

Clearer speech gives recognition a better starting point. Music, distortion and overlapping speakers can interfere; noise reduction will not reconstruct every missing word. Plan on an internet connection for AI processing and account checks.

Translate captions or keep two languages on screen

There are two documented workflows, and the difference matters if you have already corrected the text:

• Translate existing captions, when that option is present. CapCut's video-to-text instructions describe generating and editing captions, then opening Translation. Select the reviewed captions, choose the target language and compare the translated lines with your approved source.
• Generate bilingual captions from the audio. The bilingual-subtitle instructions describe recognition and translation together. Desktop uses Captions → Auto captions; Mobile adds a target language under Translate to. A new Generate action can recognize the speech again and bring back errors you already fixed.

Before generating again, preserve the reviewed track in your separate copy and inspect any replace/delete-existing-captions option. Hiding a track is not protection against replacement. After generation, compare both new language tracks with the recording and approved text.

For two visible lines, platform-specific controls and phone layout, follow the bilingual subtitles guide. Check the available source and target languages separately; a language offered for captions is not necessarily available for dubbed speech.

The source-first order in this diagram is a review principle. It does not mean that every Generate button translates an edited transcript. If your version has no existing-caption Translation option, correct the resulting target text manually or use a reviewed subtitle file.

How to translate spoken audio with AI dubbing

CapCut's AI Dubbing instructions describe Desktop and explicitly restrict availability to certain regions. Its detailed dubbing walkthrough places the control in the selected clip's right-hand Audio → Basic settings—not the music library.

1. Open your short test project. Put the speech video on the timeline and select it. Keep the original recording available for comparison.
2. Inspect the clip's right-hand Audio panel. In the documented layout, open Basic and look for Video translator. If it is absent, use the availability checks below; don't search indefinitely through music assets.
3. Choose the source and target languages. Confirm what the speaker actually says rather than relying only on automatic detection. Use a target offered by this specific tool.
4. Read the generation prompt. Check any subscription, credit, clip-length or regional restriction before starting. A free-download button on a website does not establish this action's price.
5. Generate the sample. Wait for processing to finish. Don't assume that corrections in a separate caption track control the dubbed wording.
6. Compare the result with the original. Check meaning, names, numbers, pronunciation, pauses, volume and any processed mouth movement. Export and review the sample before continuing with the full project.

The diagram's “Live gate” and “Generate one test” are explanatory labels, not buttons to find in CapCut. Use the actual prompt in your editor. If a dubbed phrase is wrong and the tool offers no way to edit its script, a reviewed voiceover or narration made from approved text gives you more control.

Review the translated video in the right order

Ask someone fluent in the target language to review anything you cannot confidently judge yourself. Translating the output back is a useful warning check, not proof that it sounds natural or preserves the meaning.

1. Meaning: compare the recording, available transcript and translated result. Check who said what, including negatives and uncertain statements.
2. Names and numbers: review names, brands, dates, prices, measurements and abbreviations individually. Confirm pronunciation as well as spelling.
3. Caption timing: allow enough time to read each line. Recheck long translations, line breaks and text near app controls at phone size.
4. Voice: listen for clipped words, rushed endings, awkward pauses and changes in tone or loudness between clips.
5. Lip-sync: inspect close shots and cuts. Look for speech starting before the mouth opens, movement after speech ends, or a visibly mismatched closure. No short sound checklist proves every language or scene is correct.
6. Export: play the finished file outside CapCut, with the final captions and audio together. Keep the approved project and source files.

Which CapCut platforms support video translation?

Use the feature-availability checks before reinstalling. An official landing page describes a product capability, not what every signed-in account can use today.

Is CapCut Video Translator free or Pro?

CapCut promotes free translation on its product pages, but access can depend on the action and account. Its AI-access checklist distinguishes regional restrictions, Pro requirements and credits. It does not establish one universal dubbing allowance.

Check the prompt before generation and again before export. If a charge or quota appears, note what it applies to. A Pro subscription does not override a region restriction, and a free caption action does not prove dubbed audio is free. Test the workflow you need before paying for a long project.

CapCut Video Translator missing or not working

Use the diagram to identify the problem, then follow the qualifications below. “Regenerate” is not a universal repair: the next run may start from the audio again rather than your corrected text.

Avoid repeating the same translation mistake

Changing the app language changes menus, not the speech. Editing a caption's spelling does not automatically repair a dub. Generating again from unchanged audio can repeat the same recognition error. Keep the original available until the full translated result—not just one attractive preview—passes review.

For basic transcript correction and styling, use Auto Captions. Voice effects and custom-voice options are separate tasks in the AI Voice guide.

Use an approved translation when automation falls short

For subtitles, prepare a reviewed SRT with cue numbers and timecodes intact, then follow the subtitle import guide for documented Web/Desktop routes and Mobile qualifications. Translate the text, not the timing syntax. A renamed plain TXT transcript does not acquire timing.

For speech, record approved narration or generate it from reviewed text, then align it with the edit. That is not the same as cloning the original voice or automatically matching the mouth. If the client needs a subtitle file rather than text burned into the video, use the caption-file export guide and check the exported cues. Burned-in text cannot be switched off by the viewer.

Privacy, rights and consent

Get permission to use the source material and any recognizable voice or likeness. Keep approved scripts, and don't make somebody appear to endorse a statement they did not approve.

CapCut's AI safety guidance advises caution with confidential files and other people's face or voice. For client work, confirm that cloud processing is allowed before uploading. Processing, exporting and publishing are different actions; none is a blanket privacy guarantee. Remove private screens, addresses and account details from test material.

CapCut Video Translator FAQ

Before you translate the next video

Keep the approved source, review both text and voice, and watch the exported file before repeating the workflow. Use the AI tools hub for related workflows and export settings for resolution, frame rate and codec checks.