Modules & lessons
Course & training teams
The same lesson goes out in six languages, and the captions have to look like your course, not a default.
Subtitles & Translation
AI subtitles: generate and translate in 95+ languages.
Subtitle a video once, translate it into as many languages as you need, and style the captions your way. Download SRT, VTT or FCPXML, or the video with subtitles burned in.
Start for FreeFree 15 minutes · no credit card
plus.subanana.com
Star Ferry crossing
Proofread
It still costs less than a bottle of water.Cela coûte encore moins qu’une bouteille d’eau.
00:19 / 08:42
Timing🇬🇧 English · source🇫🇷 FrenchCPS
00:12.400:15.1The Star Ferry has crossed this harbour since 1888.Le Star Ferry traverse ce port depuis 1888.14
00:19.000:22.6It still costs less than a bottle of water.Cela coûte encore moins qu’une bouteille d’eau.17
00:27.200:29.9Sit on the upper deck if you want the view.Installez-vous sur le pont supérieur pour la vue.13
If you publish the same video in more than one language, this is for you.
Four teams, four kinds of video, one problem — the second language costs more than the first.
Episodes & shorts
Podcast & video creators
You want the subtitle file, the video with captions burned in, and the description written.
Campaigns & films
Marketing & brand
Product names have to survive the translation, and the caption has to sit inside the brand's type.
Interviews & archive
Media & publishers
You deliver bilingual subtitle files to a house spec, and somebody downstream has to conform them.
How does it work?
Star Ferry crossing
MP4 · 08:42 · uploaded
Upload a file, paste a link, or recordor switch an existing project to subtitles
Subtitles
00:12
Let's start with the second module.
00:19
Open the file you exported last week.
00:27
The settings panel is on the right.
north windNorthwindglossary
Translated in one pass
🇫🇷 FR🇪🇸 ES🇩🇪 DE🇵🇹 PT🇯🇵 JA🇰🇷 KOBurned-in video
MP4 · 720P to 4K
Subtitle files
SRT · VTT · FCPXML
Bilingual tracks
primary + secondary
Text and data
DOCX · XLSX · TXT · Markdown
Ask the video
descriptions, chapters, quotes
Nine reasons this beats subtitling it twice.
Not a feature list — the things that decide whether the second language costs you a day or a click.
Bring it in from anywhere
- Upload or paste a linkMP4, MOV, M4A or MP3 — or a YouTube, Instagram or Facebook URL.
- Record in the browserCapture the screen or your voice with no other software installed.
- Bring subtitles you haveImport an .srt or .vtt into a project and translate that.
- Reuse a transcriptSwitch a finished transcription to subtitles, at no extra cost.
Every language in one pass
- Pick them all up frontChoose as many targets as you need before processing starts. 95+ to pick from.
- One track per languageEach becomes its own subtitle track you can open, edit and export on its own.
- Redo one languageRegenerate or delete a single translation track; the others stay as they are.
- Add a language laterOnly the new track is processed; the finished ones stay untouched.
Your names survive translation
- Mark a term universalIt comes back spelled that way in every track, not translated into something else.
- Or pin it per languageWhen a market really does use another name, set that rendering just for it.
- A workspace glossaryApplies automatically by language, on every project you start.
- Import or extractLoad terms from CSV or XLSX, or let the AI pull candidates out of a PDF.
Captions that look like you
- Two caption stylesOutline and shadow, or a background box with its own opacity.
- Styled per trackPrimary and secondary styled separately, so bilingual subtitles stay readable.
- Adaptive font sizeLong lines stay inside the frame instead of running off the edge.
- Previewed on the frameFont, size, colour, alignment and margin, shown on the real still as you set them.
Burn it in, up to 4K
- Up to 4K720P, 1080P, 2K or 4K, with the subtitle style exactly as you set it.
- Bilingual burn-inPrimary and secondary stacked on one frame, each styled separately.
- Your watermark, or noneHide the default one entirely, or replace it with your own.
- Renders in the backgroundKeep editing; the job uses a snapshot from when you pressed render.
Proofread before anyone reads
- A score out of 100Suggestions sorted by kind, so you review them in groups.
- Approve by groupApply the groups you agree with; everything else is left exactly as it was.
- Timings never moveMisheard words and punctuation get fixed without the cues shifting underneath.
- Entirely optionalNothing is applied until you press the button. Skip it if you would rather.
Chapters and copy come with it
- Chapter markersReal timestamps, ready to paste under the upload.
- Description and tagsSEO tags, the best pull quotes, or a blog draft off the same transcript.
- Translated summariesThe second market gets more than translated captions.
- Ask about the videoAsk a question about the transcript and get an answer from what was said.
Invented lines get caught before export
- Checked automaticallyEvery transcript is checked for text that does not match the audio.
- Automatically re-runA flagged section is redone on another AI model before it reaches your file.
- Silence stays silentA pause is a pause, not a sentence the model imagined.
The file the platform asks for
- SRT, VTT or FCPXMLOne file per language, timed to the cut you approved.
- Readable formatsTXT, DOCX or Markdown when somebody needs to read it, not play it.
- One ZIPThe original and every translation draft inside, for handing to a reviewer.
- Timestamps your wayIn or out, with a blank line between cues if that is the format you need.
And the rest of it.
None of these are the reason to switch, but every one of them is there when the delivery spec asks for it.
Upload a filePaste a linkScreen recordingVoice recordingImport an .srt or .vttFind and replaceTimeline with soundwavesCustom watermarkExport historySRT exportVTT exportDOCX exportXLSX exportCustom spreadsheet columnsTXT exportMarkdown exportShare linkWorkspace rolesSwitch project type
The detail lives in the product — a free workspace takes about a minute to set up.
See it in the productQuestions people askbefore they hand over a video
An uploaded file, a link, a screen recording, or a voice recording made in the browser. If you already have a subtitle file, upload the .srt or .vtt and translate from that.
Pick as many as you need before processing — 95+ are available, including Cantonese. The free plan runs up to three at a time; paid plans are unlimited.
Yes. Font, size, text, background and outline colour, outline-and-shadow or background-box, alignment, bottom margin, opacity and adaptive sizing — set per track, so a bilingual pair can be styled differently.
Yes, as a real render at 720P, 1080P, 2K or 4K, single or bilingual, with the default watermark hidden or replaced by yours. The free plan renders a 15-second clip.
Before processing: put it in the glossary, or attach a background pack so the engine reads your script first. After: fix it once with find and replace across every cue, then run an AI proofread over the result.
It is not a video editor — no cuts, no colour, no dubbing or synthetic voice. And it will not rescue a bad recording: a phone on a table in a loud room is still a phone on a table in a loud room.
Questions people askbefore they hand over a video
An uploaded file, a link, a screen recording, or a voice recording made in the browser. If you already have a subtitle file, upload the .srt or .vtt and translate from that.
Pick as many as you need before processing — 95+ are available, including Cantonese. The free plan runs up to three at a time; paid plans are unlimited.
Yes. Font, size, text, background and outline colour, outline-and-shadow or background-box, alignment, bottom margin, opacity and adaptive sizing — set per track, so a bilingual pair can be styled differently.
Yes, as a real render at 720P, 1080P, 2K or 4K, single or bilingual, with the default watermark hidden or replaced by yours. The free plan renders a 15-second clip.
Before processing: put it in the glossary, or attach a background pack so the engine reads your script first. After: fix it once with find and replace across every cue, then run an AI proofread over the result.
It is not a video editor — no cuts, no colour, no dubbing or synthetic voice. And it will not rescue a bad recording: a phone on a table in a loud room is still a phone on a table in a loud room.