Multilingual Transcription: Foreign-Language Video (2027)

Summary

For multilingual transcription, transcribe in the language that was spoken, correct that transcript, then translate. One corrected source means one set of fixes instead of one per language.

Set the spoken language yourself when you know it. Auto-detect is less reliable on accented speech and on recordings that mix languages.

VideoText's Video to Transcript transcribes in 104 languages, and Translate Subtitles covers 104 target languages with every cue's original timing kept.

Transcribe your foreign-language video in Video to Transcript, free for files up to 30 minutes.

Transcribing a foreign-language or multilingual video goes wrong in predictable ways: it's transcribed in the wrong language, the translation drifts out of sync, or a non-Latin script renders badly once it's burned in. Each is avoidable if you do the steps in the right order.

This guide is for creators publishing in more than one language, teams localizing video, and anyone who needs text from a recording that isn't in English.

Start now: upload the recording to Video to Transcript and set the spoken language. Free account, no card needed.

The order that works

Step	What you do	Why this order
1	Transcribe in the spoken language	Every later file is built from this one
2	Correct the source transcript	One correction fixes every translation made from it
3	Make subtitles from the same pass, if you need them	Timing comes from the audio, not from a translation
4	Translate the transcript or subtitle file	Timing carries over; only the text changes
5	Fix readability per language, then style or burn	Translations run longer, and scripts render differently

Step 1: Transcribe in the language that was spoken

On Video to Transcript, pick the spoken language from the transcription language picker, which lists 104 languages, or leave it on auto-detect. When you know the language, set it. VideoText's own guidance is that auto-detect is less accurate, particularly for accented English and mixed-language recordings.

Add names and terms before you start: product names, people, places and jargon are where any transcript is most likely to be wrong, and they're the words a viewer notices.

Free handles files up to 30 minutes and 2 GB. Pro handles files up to 4 hours and 10 GB, so a full lecture, panel or interview goes in as one file with no joins to check.

Step 2: Correct the source before you translate

Every error in the source transcript is copied into every language you translate into. Fix it once, at the source. If a native speaker is available, this is where their time is best spent.

For a transcript you already have, Fix Transcript suggests fixes for grammar, punctuation, capitalization, likely mishearings, times and paragraph breaks, and you accept or reject each one.

Mixed-language recordings. If speakers switch languages mid-recording, set the main spoken language and review the switched passages by hand. Don't assume any tool has handled them.

Step 3: Make subtitles from the same recording

If the video needs subtitles, generate them from the recording in Video to Subtitles rather than splitting a finished transcript into cues by hand. You get timed SRT and VTT files from the same cues, and you can add a translated lane with the same timing and download it as its own file.

Step 4: Translate with timing kept

Translate Subtitles takes an SRT or VTT file up to 20 MB and translates it into any of 104 target languages. Every cue keeps its original start and end time, and the format stays the same: SRT in gives SRT out, VTT in gives VTT out. Lines that look untranslated or mixed-language are pointed out, so you know where to look. It also translates TXT, DOCX or JSON documents, or pasted text.

You translate one language per run, so run it again for each language. On Pro you can correct any translated line before downloading.

For plain transcripts, Video to Transcript can translate as part of the job: choose a language before you start, or paste a transcript into "Translate an existing transcript". This works with a free account.

Our guide to translating subtitles without losing timing sync covers the timing side in more depth.

Translate your subtitle file now: open Translate Subtitles, pick one of 104 target languages, and download a file with every cue's timing unchanged.

Step 5: Fix readability per language, then style

Translations often run longer than the source, so a file that passed every check in one language can fail line length or reading speed in another. Run each translated file through Fix Subtitles, which wraps lines to 42 characters and gives cues faster than 20 characters per second more time where the next cue allows it. If your client or platform sets different limits, follow their spec.

If you burn captions into the video, check the script renders well. Subtitle Style Studio uses script-aware fallback for Arabic, Indic and CJK captions, and you approve a real rendered frame before export. Burn Subtitles can then burn each language into its own MP4.

Common mistakes

Leaving auto-detect on for a known language. Set it when you know it.
Translating before correcting. Each source error becomes several.
Re-timing translations by hand. If timing is kept, you don't need to.
Skipping the per-language check. Translated lines are longer.
Burning in before checking the script. Check a rendered frame first.

What it costs

Transcription and translation both start with a free account, no card needed. Free covers files up to 30 minutes, with a tool-specific translation allowance, and free exports carry a VideoText watermark. Pro is $14.99/month or $120/year, with 30 hours of speech a month, transcription files up to 4 hours, translation in 104 languages with timing kept, and no watermark on exports. See pricing.

FAQ

How do I transcribe a video in another language? Set the spoken language yourself, add the names and terms you expect, transcribe, and correct the transcript before you do anything else with it. Translate from that corrected transcript.

How many languages does VideoText support? 104 for transcription in Video to Transcript, and 104 target languages for subtitle translation.

Can I get subtitles in several languages from one video? Yes. Generate subtitles once, then translate the file into each language. One language per run, and every cue keeps its timing.

What if speakers switch languages during the recording? Set the main spoken language and review the switched passages by hand.

Next step

Upload your next non-English recording to Video to Transcript with the spoken language set, then translate the subtitles with timing kept. Free covers files up to 30 minutes; Pro handles files up to 4 hours, removes the watermark and lets you edit translated lines, for $14.99/month or $120/year. This guide is part of How to transcribe a video to text in 2027 and Subtitle and transcription tools: the 2027 guide.

Multilingual transcription in 2027: transcribe foreign-language video, then translate

The order that works

Step 1: Transcribe in the language that was spoken

Step 2: Correct the source before you translate

Step 3: Make subtitles from the same recording

Step 4: Translate with timing kept

Step 5: Fix readability per language, then style

Common mistakes

What it costs

FAQ

Next step

Related guides