Upload the video or paste a link. The sous-titres come back in French — the words as spoken, never an English version — timed from word-level timestamps and ready to import. Files over 500 MB continue on the upload page. A pack from $4 opens the full transcript · packs from $4 for 500 minutes.
How it works
The soundtrack is read out of the video; there is nothing to extract first.
An MP4, MOV or WebM goes into the box above, or paste a public link. Over 500 MB, continue on the upload page.
Lines are built from the timestamp carried by each word, so a cue arrives on the syllable rather than on an average. Your first transcript opens in full at any length.
SRT for an editor or a platform upload, VTT for a player on the web, and TXT, DOCX, MD and JSON from the same job.
Making them readable
Timings come out of the job. Line breaks and trims are editorial work you do in a subtitle tool.
The same scene usually needs more characters in French than in English. A cue that was roomy in the English version can arrive over the line limit, so plan on trimming a few.
Break after punctuation or before a conjunction rather than inside a noun phrase. Around 42 characters a line and two lines a cue is the common ceiling.
French typography puts a space in front of the double punctuation marks. Use a non-breaking space so a line never begins with a lone question mark, and check how your editor renders it.
Circumflexes, cedillas and the œ ligature are written into the file. Import and save as UTF-8 or they arrive as mojibake.
Sidecar or burned in
You get the sidecar file; the render into the picture happens in your editor.
SRT or VTT sits beside the video. Viewers can switch it off, platforms can read the words, and correcting a name later does not mean exporting the video again.
When subtitles have to be permanent, bring the SRT into Premiere Pro or DaVinci Resolve, style it to your house look and burn it in there.
SRT and VTT carry words and timings only. For a documentary that names its contributors on screen, add those in the editor; JSON is where the speaker of each segment is kept.
Upload the soundtrack on its own if the master is enormous. The timeline does not move, so the cues still land correctly.
Pricing
A pack from $4 opens the full transcript. After that, one-time packs: $8 for 1,000 minutes, $12 for 2,000 minutes and $24 for 5,000 minutes. Credits never expire.
A 12-minute piece is about $0.10 on the $8 pack. A 90-minute documentary is $0.72 on the same pack.
FAQ
Yes. The file holds the French that was spoken. It is not turned into English: French video produces French cues, and that is what you download.
Yes, and there is no extra charge for a second format. Every job exports SRT, VTT, TXT, DOCX, MD and JSON.
From word-level timestamps. Each word is timed against the audio and the cues are built from those times.
No. Numbered speaker labels appear in the transcript view on screen only. SRT and VTT hold words and timings; JSON is the export that keeps the speaker of each segment.
Usually. Public video links are fetched and transcribed, though some sources refuse to hand over audio — Spotify, Podbean and Kick VODs among them — and those you upload.
That stretch comes back as English text and the French stretches come back as French. The file follows the speakers; nothing is translated.
Up to 10 hours or 5 GB per file. The box here takes 500 MB, larger files go through the upload page, and your first transcript reads in full at any length.
Instant preview, no signup. A pack from $4 opens the full transcript.