Whether an MP4 has subtitles you can extract depends on whether it carries a text track. Most do not. Here is how to check in seconds, how to pull out a track when it exists, and how to generate a clean SRT from the audio when it does not.
Manual translation, hired voice actors, days of waiting. Creators and teams lose time and money on every single video.
Skip the old workflowI was told 'just extract the subtitles' as if MP4 magically contains them, and after an hour with tools I have nothing.
The subs are burned into the video and I need the text back — to translate it, to correct it, or to reuse it as a caption file.
There is a text track but it is a mov_text stream that most editors will not open, and ffmpeg flags are not my idea of a good afternoon.
Upload, pick languages, and let the AI handle the rest. No software to install.
A few clicks, that is all.
A few clicks, that is all.
A few clicks, that is all.
A few clicks, that is all.
MP4 subtitle extraction has a reputation problem: countless pages promise the trick, and the trick usually does not exist for the file you actually have. There are only two situations. If the MP4 carries a text subtitle track — in MP4 that is a mov_text (also called tx3g) stream, which is common in files from broadcast or professional workflows — you can copy it out losslessly and the video is untouched. If it does not, there is nothing to extract, and no software can find subtitles that were never in the file. That second situation is by far the more common one, because the ways MP4s get subtitles in the real world are burning them into the picture (screen recordings, social exports, most downloaded videos) or shipping an external .srt alongside. When there is no track, the workable route is to build one: transcribe the audio and let the timing come from the speech. Upload the MP4 here and the output is a standard SRT with the original timeline of the video, editable line by line, plus TXT if you only wanted the words. 1 credit per minute of audio, 20 free at sign-up.
Subtitle → Subtitle Track in VLC tells you which route you are on before you waste time on tools: a named track is extractable, an empty list means your only option is generating one.
Burned-in text cannot be extracted by any tool — it is part of the picture. What works is generating a fresh SRT from the audio, with timing from the speech, which is also what you need when the original subs were mistimed.
Generated subtitles arrive as normal cues: fix a name, split a long line, nudge a timing. Extracted tracks often carry the original broken line breaks and styling you were trying to escape.
First check whether the file has a text subtitle track (VLC → Subtitle → Subtitle Track). If it does, extract it losslessly with one ffmpeg command, for example ffmpeg -i input.mp4 -map 0:s:0 output.srt — the video is untouched. If it does not, extraction is impossible and the subtitle text has to be generated from the audio, which is what this tool does.
Because most MP4s never had one. Soft subtitle tracks in MP4 use the mov_text format and come mainly from broadcast, streaming or professional authoring pipelines. Files from screen recorders, phones and social downloads either have the subtitles burned into the frames or shipped as a separate .srt beside the video.
Not by extraction — the text is made of pixels, so there is no track to copy. You can generate a replacement: transcribe the audio and produce a new SRT with timing taken from the speech. That is also the usual fix when the original subtitles were badly timed, in the wrong language, or simply ugly.
mov_text (also called tx3g) is the timed-text format MP4 and MOV use for soft subtitles. It is plain text with timing, similar in spirit to SRT but packaged as a stream inside the container, which is why some editors and converters refuse to open it directly and you need -map to pull it out.
Yes. Timestamps are generated from the speech inside the audio, so the SRT lines up with the video you uploaded rather than with an arbitrary start point. If the audio starts with a silent leader or a title card, the first cue simply starts later — nothing needs shifting.
Import it into Premiere, CapCut, Final Cut or DaVinci, upload it to YouTube alongside the video, or keep it as your master copy. You can also translate it into 67+ languages with the timeline preserved — and if the video already has hardcoded subtitles in the old language, erase them and burn in the translated set instead.
We use local storage only to keep you signed in — no tracking cookies, no third-party ads. See Privacy Policy
SpeakVid does not use tracking or advertising cookies. The only storage used is listed below.