Whether you can extract subtitles from an MKV depends on one thing: does it have a soft subtitle track? Here is how to check, how to extract one, and how to generate an SRT from the audio when there is nothing to extract.
Manual translation, hired voice actors, days of waiting. Creators and teams lose time and money on every single video.
Skip the old workflowEvery guide I find tells me to install ffmpeg and type command-line flags. I just want the subtitles out.
The MKV is soft-subtitled, but the subtitle track is ASS with styling and the player keeps overriding my font and size settings.
My file has subtitles burned into the picture and no track at all — some tools still pretend they can 'extract' them.
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.
MKV is a container, not a video format, and that is the whole story. A single .mkv file can hold a video track, several audio tracks and any number of subtitle tracks — SRT or ASS as text, or PGS as pictures — and players switch between them on the fly. So "extract subtitles from MKV" is really a question about your specific file, and there are only two answers. If a subtitle track exists, extraction is a fast, lossless copy: it lifts the text out and leaves the video untouched. If no track exists — burned-in subtitles that are part of the picture, or a file that never had them — there is nothing to extract, and the only way to get subtitles is to transcribe the audio and build the timings from speech. SpeakVid handles that second route: upload the MKV (or the audio pulled from it), and you get a standard SRT with the timeline generated from the speech itself, editable line by line before export. 1 credit per minute of audio, 20 free at sign-up.
Extraction is a copy and only works when a track exists; generation is transcription and works on anything. Knowing which one your file needs is the entire problem — and it can be checked in ten seconds.
Text that is part of the picture cannot be extracted by anything — no tool can un-burn it. What you can do is generate a fresh, clean SRT from the audio, which is also what you need if the original subs were the wrong language or badly timed.
Generated subtitles arrive editable: correct a name, split a cue, adjust a timing. Extracted tracks usually come back with the same broken line breaks you were trying to escape.
Because extraction is a copy, and there has to be something to copy. If the file has no subtitle track — subtitles burned into the picture, or a video that never had any — nothing can pull subtitles out of it. In that case the only route is transcription: build the subtitles from the audio, which is what this tool does. Files that do have a soft track can be extracted losslessly with a one-line ffmpeg or mkvextract command.
Open it in VLC or MPV and check Subtitle → Subtitle Track. A popup with named tracks (for example English, Forced, Signs) means there is something to extract and you can also just play that track inside the player. An empty menu, or subtitles that stay visible when subtitles are switched off, means they are burned into the video.
SRT carries plain text plus timecodes, ASS adds styling and positioning, and PGS is bitmap images rather than text — that is why PGS files cannot be edited as text and are slow to re-style. When you generate subtitles from the audio here you get a standard SRT, which every player and editor reads and which stays fully editable.
No — extraction is a local file operation, and the honest answer is that a one-line command does it better than any web tool. What SpeakVid does is the route that works when extraction is impossible: transcribing the audio into a new, cleanly timed SRT. Many people also prefer this route even when a track exists, because the original subtitles are often badly timed or in the wrong language.
Up to 98% on clean audio, with automatic sentence breaks and punctuation, and timestamps on every line so you can check a passage by clicking it. Accuracy falls with background music, overlapping speakers and muffled dialogue. You can correct any line in the editor before exporting.
Export SRT to import into Premiere, CapCut, Final Cut or DaVinci, or to upload alongside a video. Export TXT if you only wanted the text. You can also translate the SRT into 67+ languages with the timeline preserved, or use the video tool to erase burned-in subtitles and burn in a new set — that is the usual move when you have translated the file and want the old text gone.
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.