01 / INSIDE THE MP4
An MP4 is read for the audio inside it.
An MP4 file carries both picture and sound. For a transcript, the picture does not matter: the mp4 to transcript page reads the audio track inside the file and works from that. You do not export the sound separately first, and you do not need a video editor to prepare anything. Choose the MP4 as it is.
That has a practical consequence worth stating plainly. This page reads a file you already have, not a link and not a caption track pulled from a website. If the only copy of the recording lives at the end of a URL, this is the wrong page; a link belongs on the YouTube page instead. If the recording is sitting in your downloads folder, on a phone, or on a drive from a camera, this is exactly the page for it.
It also explains why the page works when a video has no captions at all. Nothing is being looked up and nothing is being read off the video. The words come from the sound, which is present in every recording that has speech in it.
02 / FORMATS AND LIMITS
Formats, and the two limits that matter.
The file control names six formats: MP3, M4A, WAV, MP4, MOV and WebM. The last three are video containers and the first three are audio files, which makes the page useful for a voice memo as well as a camera clip. A file the browser cannot read comes back with a message suggesting MP3, M4A or WAV, so there is no guessing about what to try next.
Two limits are enforced rather than advised: 20 minutes and 200 MB. The size limit is checked in the browser the moment you choose a file, and a file over the cap is refused with a plain sentence rather than a stalled progress bar. The length limit is checked on the way through, and a recording past it fails instead of being quietly trimmed.
For a longer recording, the workable approach is to split it into parts that each sit inside twenty minutes and transcribe them one after another. Each part is a separate pass, and the text still ends up in the same three files.
03 / THE STEPS
The three steps, in the order they happen.
Choose the file.
Use the button at the top of the page. The browser opens your own file system and sends short pieces of the audio, not the whole recording at once.
Wait while the audio is worked through.
The browser splits the audio into short pieces and sends each piece for recognition, then the words are put back in order. A longer file simply takes longer; the status line reports progress.
Copy or download.
Copy takes the text as it stands. Downloads save TXT for reading, SRT for a subtitle workflow, or VTT for a web player.
Note the order of those steps if you are troubleshooting. Nothing is decided at upload time except whether the file can be read and whether it fits the limits. Any problem with the words themselves appears in step two, not step one.
04 / WHAT YOU GET
Three files from one transcript.
Every successful run ends in the same result, and the same three downloads plus a copy button.
- Copy puts the text on the clipboard and creates no file at all. It is the quickest way into a document, a notes app or a message.
- TXT is the words on their own: the reading copy, with no clock anywhere in it.
- SRT puts a number and a clock on each cue, the format a subtitle workflow expects.
- VTT carries the same cues for a web player.
One wrinkle is worth knowing before you open an editing program: when a result arrives with no timings, SRT and VTT stay switched off and only the text is available. That is not an error. It means there was nothing to time, so the page gives you what it actually has instead of a file full of zeros.
Once you have the text, the question becomes how much of it to trust. Names, numbers and unusual words are where speech recognition tends to slip, and the article on getting a more accurate transcript covers what you can do about that, most of it before the file is ever chosen.
05 / LANGUAGE AND EDITS
An English page that leaves the recording as it is.
The site and its interface are in English. The recording is not forced into English: the speech request does not set a language, so a file is transcribed in the language actually spoken in it. That distinction matters for anyone working with a mix of material, because an English page is not the same as an English-only listener.
What the page will not do is edit. Nothing in the words is reworded, no timing is shifted, and no caption is burned into the picture. Those are separate jobs with separate tools, and this one stops at the text.
Where the file goes is worth a sentence too. Your browser cuts the sound into short pieces and hands each one to the site; the server passes it on for recognition and returns the text. The download file is built on your own device.
06 / QUESTIONS
Do I need to extract the audio first?
No. Choose the MP4 as it is; the browser reads the audio inside the file.
What are the size and length limits?
Up to 20 minutes and 200 MB per file. A file that is longer or larger is refused.
Can I use an audio file instead of a video?
Yes. MP3, M4A and WAV are named alongside MP4, MOV and WebM.
Does the page edit the words or add the captions to the video?
No. It returns the text as it was recognised. No word is reworded, no timing is shifted, and no caption is burned into the picture.
READY WHEN YOU ARE
Bring the file you already have.
No link and no captions needed. Choose the MP4 and take the words away as text.
Upload an MP4 and get the text ↗