Checked August 28, 2026. CapCut currently uses “video translator” for two different jobs: translated subtitles and translated speech. The subtitle route is documented on Web, Desktop and Mobile; the AI-dubbing route is advertised mainly through Desktop and is explicitly limited in some regions. Menu labels, languages and plan or credit gates can differ by account.
If you only need viewers to read another language, generate and translate captions. If you need them to hear another language, use AI dubbing when the Video translator control appears. Do not judge either result from the generate button. Correct the source transcript, names, numbers, timing and pronunciation, then watch the exported file outside CapCut.
Choose subtitles or AI dubbing
Translated subtitles
Use this route when the original voice can stay and viewers need captions in another language. The caption blocks remain editable, so you can correct terminology, timing and line breaks.
AI dubbing
Use this route when the spoken track itself must change language. Review meaning, voice continuity, pronunciation and mouth timing before you trust the full export.
Dubbed audio plus captions
Generate and review the audio and text as separate layers. A correct subtitle track does not prove that the dubbed voice is correct, and the reverse is also true.
Approved translation or voiceover
Use a reviewed script, imported subtitle file or recorded voiceover when names, legal wording, medical language or brand terms cannot be left to an automatic first pass.
| You need | Start with | What changes | Main review risk |
|---|---|---|---|
| Another-language text | Captions / Auto captions | Caption blocks | Meaning, spelling, timing and readability |
| Another-language speech | Audio / Video translator when shown | Spoken audio; lip movement may also be processed | Meaning, pronunciation, pacing, voice continuity and lip-sync |
| Both audio and text | Dubbing, then captions | Two separate output layers | Text and speech disagreeing after edits |
| Controlled terminology | Reviewed script or subtitle file | Only the approved text/audio you supply | Import support, timing and version control |
What CapCut Video Translator actually does
CapCut’s naming is the confusing part. Its official Video Translator page describes speech recognition followed by translated subtitles. Its separate AI Video Translator page describes translated speech, voice-style preservation and lip-sync processing. The two pages target the same broad query, but the outputs are not interchangeable.
Caption translation follows a text chain: speech becomes a transcript, the transcript is translated, and the translated caption blocks are placed on the timeline. A bad source transcript produces a bad translation even when the translation engine does exactly what it was asked to do. Fix “Maya” being transcribed as “my year” before translating it into five languages.
AI dubbing adds more failure points. It must identify speech, translate the meaning, generate a target-language voice, fit different sentence lengths into the scene and, where offered, align the visible mouth. CapCut markets natural voice retention and lip-sync, but those are intended capabilities, not a promise that every face, accent, cut or overlapping speaker will pass.
Before you translate a real project
- Duplicate or back up the project. Keep the original speech, source media and an untouched project copy. Do not make the only live edit your translator test.
- Choose a representative short clip. Include one name, one number, a normal sentence and a visible face if lip-sync matters. A clean ten- or twenty-second test reveals more than generating an entire interview first.
- Reduce avoidable audio problems. Loud music, echo, clipped speech and people talking over each other weaken recognition. If you cannot clean the track, expect more manual correction.
- Write a terminology sheet. List names, brand spellings, acronyms, prices, dates, units and words that should not be translated. Decide whether a person’s name stays unchanged or receives a target-language pronunciation.
- Confirm rights and consent. Make sure you may use the source footage and create a translated or voice-preserving version. CapCut’s commercial-use and materials guidance does not turn an unlicensed clip or an unauthorized voice into cleared content.
How to translate subtitles in CapCut
This is the safer starting route for most projects because the original voice remains audible and every translated caption can be edited. CapCut’s current bilingual-subtitles help page documents routes for Web, Desktop and Mobile, but the placement and wording are not identical.
- Import the source video. Put the clip on the timeline and confirm the spoken track is audible.
- Open Captions or Auto captions. Use the label shown on your platform rather than searching for an Audio dubbing control.
- Select the language actually spoken. Do not select the language you want to create as the recognition language.
- Generate the source transcript first. Read it before translation. Correct names, numbers, punctuation and missing words.
- Choose the target language and translate. Keep the source captions until you have compared both versions.
- Review timing and export. Check line breaks, safe area and reading speed, then watch the exported file on another device.
Web
Open the Web editor, add the video, then use the Captions area to recognize speech and translate the generated text. CapCut’s current Web resource also describes a Translate control near the subtitle workflow. If you do not see the same wording, look for the translation action inside Captions rather than changing the website interface language.
Desktop
Open Captions, choose Auto captions, select the spoken language and the target language offered, then generate. The exact layout can vary by build. If you already have a professionally translated SRT, the official subtitle-import guide documents Desktop and Web import routes, which may be better than retranslating approved copy.
Mobile
Open the project, tap Captions, then Auto captions. Select the recognition language and, when shown, the Translate to language. Mobile does not currently offer the same direct external subtitle-file import described for Desktop and Web, so do not build a mobile-only workflow around importing SRT files.
How to translate spoken audio with AI dubbing
CapCut’s current AI Dubbing page advertises the Desktop route as Audio → Video translator after a clip is on the timeline. The page also says the feature is available only in specific regions. Treat that note as a hard availability warning, not small print.
- Select one short timeline clip. Start with the representative test you prepared, not the entire project.
- Open Audio and find Video translator. If it is absent, do not assume the app is broken; check platform, region, account, version and rollout first.
- Choose source and target languages. Confirm the source rather than relying blindly on automatic detection.
- Read the live action gate. Note any plan, credit, duration or availability message shown for this account and this clip. Do not infer today’s allowance from an old screenshot.
- Generate one test. Keep the original audio available for comparison and avoid batching more clips yet.
- Review the result before continuing. Compare meaning, names, numbers, pronunciation, timing, volume, voice continuity and mouth movement.
Review the translated video in the right order
Do not polish caption colors while the translation is wrong. Review from the earliest dependency to the latest output:
- Source transcript: Did CapCut hear every word and assign it to the right speaker?
- Meaning: Does the target version preserve the point, tone, negation and level of certainty?
- Names and numbers: Check people, places, brands, URLs, dates, prices, measurements and abbreviations line by line.
- Timing: Are subtitles readable before they disappear? Does dubbed speech fit the cut without rushed endings or dead air?
- Voice: Listen for strange stress, pronunciation changes, clipped syllables and a voice that shifts between adjacent clips.
- Lip-sync: Watch close shots around P, B, M, F, V, S, T, D and L sounds. Cuts, hands over the mouth and profile angles can hide or expose problems differently.
- Export: Watch the actual file on another screen with headphones and speakers. The editor preview is not the deliverable.
Which CapCut platforms support video translation?
| Platform | Translated subtitles | AI dubbing | What to verify live |
|---|---|---|---|
| Web | Official bilingual-caption guidance documents a Captions route | Do not assume the Desktop dubbing control exists | Signed-in Captions tools, target languages and export route |
| Windows / macOS | Official guidance documents Captions → Auto captions | Official product pages advertise Audio → Video translator | Version, account, region, source/target languages and live gate |
| iPhone / Android | Official guidance documents Captions → Auto captions → Translate to when available | Availability can differ; do not copy a Desktop path blindly | Current bottom menu, language selector and account rollout |
If a control is missing on one platform, use the CapCut feature-availability guide before reinstalling. A missing Desktop card does not prove the whole service has been removed, and an official landing page does not prove every account can use the feature today.
Is CapCut Video Translator free or Pro?
CapCut’s public product pages use “free” prominently, but they also state that AI dubbing is region-specific, and CapCut can change plan, account and credit gates. There is no reliable reason to publish one fixed universal allowance here. Configure the real action, then read what the signed-in interface says it will consume or require.
Before generating, record four things: the platform and app version, the signed-in account, the plan or credit badge beside the action, and any limit shown for the selected clip. If the interface shows no cost or plan message, that is evidence only for that account and moment—not a promise for every reader.
CapCut Video Translator missing or not working
Tool is missing
Confirm that you are using the relevant workflow: Captions for translated text, or Desktop Audio for the advertised dubbing route. Update CapCut, sign in again, check the Web route and compare region/account availability. Do not repeatedly reinstall over an unsaved project.
No speech is detected
Play the source track, unmute it, choose the correct spoken language and test a clean section. CapCut’s Auto Captions troubleshooting guidance names unclear audio, wrong language, unstable network and outdated software among common causes.
Translation is wrong
Return to the source transcript. Fix recognition errors, split run-on sentences and correct the terminology list before translating again. Regenerating from the same bad transcript usually reproduces the same problem.
Dub is out of sync
Test a shorter clip with fewer cuts and one visible speaker. If the dubbed pacing still fails, keep translated subtitles or use a reviewed human voiceover instead of forcing the synthetic track.
Common mistakes that make localization look cheap
- Translating captions before correcting transcription. The translation inherits every recognition error.
- Changing the app language. That changes menus, not the language spoken in the video.
- Using one style for every language. Text length changes; recheck line breaks, safe areas and reading speed.
- Leaving names to automatic pronunciation. Decide the spelling and spoken form for each target language.
- Muting the original too early. Keep it available until the dubbed meaning and timing have passed review.
- Batching before a real export passes. One preview can hide a problem that repeats across every localized version.
For the original-language transcript and styling work, use the focused CapCut Auto Captions guide. If you already have translated copy and only need a synthetic narration, the Text to Speech workflow is a clearer owner. Voice changing and custom voice belong in the separate CapCut AI Voice guide.
Privacy, rights and consent
A translated subtitle track does not clear the source footage, music or graphics. AI dubbing can also reproduce or approximate a person’s voice and alter how they appear to speak. Get permission appropriate to the project, keep records of approved scripts and do not use the workflow to impersonate someone or change their meaning.
For client or confidential material, decide whether uploading the source to a cloud-processed feature is acceptable before you begin. Remove private screens, addresses, account details and unreleased footage from the test whenever possible. Preserve the original project and source recordings so an approved human translation or voiceover can replace the generated version.
CapCut Video Translator FAQ
Does CapCut translate subtitles or spoken audio?
It can do either through different routes. Use Captions or Auto captions for translated text. Use Video translator under Audio on the advertised Desktop dubbing route when that control is available.
Can CapCut create bilingual subtitles?
Yes. CapCut’s current help page documents bilingual-subtitle generation on Web, Desktop and Mobile. The menu labels and available target languages can differ by platform, version, region and account.
Why is Video translator missing in CapCut?
First separate translated captions from AI dubbing. Then check the correct platform route, current version, signed-in account, region and rollout. CapCut’s dubbing pages explicitly say the feature is available only in specific regions.
Is CapCut AI dubbing free?
CapCut promotes free use on public pages, but availability and any plan, credit or clip limit can change. Read the live signed-in action gate for the exact account and clip instead of relying on an old allowance.
Does CapCut AI dubbing keep the original voice?
CapCut advertises preservation of speaker style, but treat that as the intended feature, not guaranteed identity or quality. Compare voice continuity, pronunciation, tone and pacing against the original.
How do I fix a bad translated subtitle?
Correct the source transcript first, then regenerate or manually edit the translation. Check names, numbers, negation, context, line breaks and timing before export.
What should I do when the dubbed audio is out of sync?
Test a shorter clip with one speaker, cleaner speech and fewer cuts. Review pauses and sentence length. If it still fails, use translated subtitles or an approved recorded voiceover.
The short version
Choose the output, not the biggest AI label. Translate captions when viewers only need another-language text. Use AI dubbing when the spoken track must change and the feature is actually available. In both cases, fix the source transcript, test one short clip, review names and meaning, and watch the exported file before you localize the rest of the project.
The wider CapCut AI tools hub explains where translation sits beside captions, voice, avatars and generation. For output resolution, frame rate and codec checks after localization, finish with the CapCut export settings guide.