CapCut Video Translation: What It Does — and What It Doesn't (July 2026)

Let's be fair first — CapCut is a genuinely good product, and for many creators its translator is exactly enough:

• It's free to use (with a CapCut account) and it lives inside a full editor, so you can translate captions and keep editing in one place.
• 27+ subtitle languages, with an accuracy claim of 98% on its official pages.
• Built for social-first video — TikTok and Shorts creators can caption, translate, style, and publish without leaving the ecosystem.
• Backed by ByteDance, with a mature template library, multi-device apps, and cloud import.

(All CapCut details from capcut.com's public tool pages, verified 2026-07-03. Note this describes CapCut's international version — Jianying, the Chinese edition, is a separate product.)

Here's the boundary that surprises people searching for "CapCut video translation":

1. No dubbing in the translator. The video translator produces translated captions. Your viewers read English; they still hear Chinese (or whatever the source language is).
2. TTS and voice cloning are separate tools. CapCut has a text-to-speech feature (200+ voices) and voice cloning in its Pro tier — but they're standalone features, not connected to the translation workflow. There's no supported path of "translate my video's speech and speak it in the new language" as one pipeline. You'd be manually stitching: export subtitles → paste text into TTS line by line → place audio clips by hand → mute original audio where needed. For a 30-second clip, tedious. For a 30-minute video, a non-starter.
3. No per-sentence dub workflow. Because there's no dub, there's naturally no transcript review → translation review → voice matching → per-line regeneration flow either.

None of this is a criticism of CapCut's scope — it's a video editor with a subtitle translator, and it's honest about that. The mismatch is in the search query: people looking for "video translation" often mean the voice, and that's a different product category.

An honest decision guide:

Translated subtitles are enough when…

• Your content lives in feeds where most viewers watch muted anyway.
• Clips are short and visual; the voice is secondary.
• You want zero cost and you're already editing in CapCut.

You need dubbing when…

• Watch time matters. Long-form viewers on YouTube rarely stay through 40 minutes of reading subtitles over foreign-language audio.
• The content is the speech — courses, tutorials, interviews, commentary.
• You're building a channel in the target language, where the audience expects native-language audio.
• Accessibility: viewers who listen while multitasking get nothing from subtitles.

If you're in the second group, read on.

aitrans.video does one thing CapCut's translator doesn't attempt: it replaces the voice, controllably.

1. Upload — up to 3 hours per task (CapCut is built around short-form; long-form is exactly where subtitle-reading fatigue hits hardest).
2. Transcript review — the Chinese (or English) speech is transcribed with speaker detection; you correct any misheard line first.
3. Translation review — edit the translated script sentence by sentence before a single word is spoken.
4. Voice selection — ~1,000 voices across three TTS engines with automatic per-speaker matching; voice cloning available.
5. Dub, check, regenerate — generate the full dub, play any line, regenerate just the ones you don't like, export.

Scope note in the other direction: aitrans.video is Chinese⇄English only, and it's not a video editor — no timelines, effects, or templates. Which is exactly why the last deliverable exists:

This is the part CapCut users tend to like most. Alongside the dubbed MP4, bilingual SRT subtitles, and clean dubbed audio track, aitrans.video exports a Jianying draft project — Jianying (剪映) being CapCut's Chinese-edition sibling, also by ByteDance. If your post-production runs through Jianying, you don't get a locked file; you get a draft you can open and keep editing — trim, re-time, add graphics — with the dub already in place.

We're not trying to replace your editor. We're the dubbing step your editor doesn't have.