Free tool

MP3 to Text

Drop in an MP3 and get the words back. Up to 90 minutes a file, free, with the text, SRT and VTT emailed to you.

Looked after by Clack, a typewriter with a sheet of paper in it.

We show the transcript here and email you the text, SRT and VTT, so you still have them tomorrow.

Free. No account. 2 files a day per person, up to 90 minutes each.

Ninety, because across 96 transcripts we measured the median episode was 47.3 minutes and 96% were 90 minutes or under.

How to turn an MP3 into text

  1. 1Drop in the MP3Up to 90 minutes and 600MB. The length is read out of the file before anything uploads.
  2. 2Say where to send itThe transcript appears here and lands in your inbox as TXT, SRT and VTT.
  3. 3Wait a few minutesAn hour of audio takes a handful of minutes. You can close the tab.
  4. 4Copy or downloadRead it here, or take the files into your editor or your notes.

Questions

Does the bitrate of my MP3 affect the transcript?
Below about 64kbps it can. Speech lives mostly under 8kHz and MP3 keeps that range well, so a 128kbps file and a 320kbps file of the same recording usually transcribe identically. What hurts is a heavily compressed file — a 32kbps mono voice memo — where the compression has already smeared consonants together. If you have the original, send the original.
Is mono worse than stereo?
No, and for speech it is often better. Two people recorded on separate channels and mixed to stereo can end up quieter in the mix than either was on its own. A clean mono file at 64–128kbps is a perfectly good thing to send.
How long can the MP3 be?
Ninety minutes, and up to 600MB. Those two limits rarely collide for MP3: ninety minutes at 128kbps is about 86MB, and even at 320kbps it is around 216MB. It is uncompressed WAV that hits the size cap, which is one good reason to send an MP3.
What if my file is longer than 90 minutes?
The tool tells you before it uploads anything, and tells you the real length it read out of the file. Split it in any editor and send the halves, or use Slice’s paid tools, which have no length limit.
Should I convert my WAV or M4A to MP3 first?
Only if it is over 600MB. The tool takes WAV, M4A, FLAC, OGG and AAC directly and does not transcribe them any worse. Converting is a way to make a big file small, not a way to make a transcript better.
Will it separate the speakers?
No. The free tool gives you the words in paragraphs. Who said what is a paid feature, and saying so here is cheaper for both of us than you finding out at the end.
What do I get back?
The transcript on screen, plus a .txt in paragraphs and .srt and .vtt subtitle files. All three arrive by email as well, so they are still yours after you close the tab.
How this compares to Otter, Rev and the rest

We read the free tiers of six other transcription tools off their own pricing pages and put them in one table — longest free file, whether they label speakers, whether you need an account, and what happens when the free part runs out. It includes where we lose, because a comparison that does not is an advertisement.

What actually makes an MP3 transcribe well

MP3 has a reputation for being the lossy one, and for music that matters. For speech it mostly does not. Human speech carries almost all of its intelligibility below 8kHz, and that is the part MP3 protects longest as the bitrate drops. A 128kbps file and a 320kbps file of the same interview will, in practice, come back the same.

Where it does go wrong

  • Very low bitrates. Under about 64kbps, consonants start smearing into each other — the difference between “fifteen” and “fifty” is exactly the kind of thing that goes first.
  • Compressing twice. A file that was already an MP3, re-exported as an MP3 at a lower bitrate, loses more than either step alone. If you still have the original recording, send that.
  • Room, not format. Far more transcripts are hurt by a phone on a table across a café than by any codec. Nothing here can undo that, and a tool that promised otherwise would be lying.

The size arithmetic

The size cap is 600MB, and for MP3 it almost never bites:

  • 90 minutes at 64kbps — about 43MB
  • 90 minutes at 128kbps — about 86MB
  • 90 minutes at 320kbps — about 216MB
  • 90 minutes as uncompressed WAV — about 950MB, which is over the cap

That last line is the one worth knowing. If you have a long WAV that will not fit, exporting it as a 128kbps MP3 makes it a tenth of the size and does not measurably change the transcript. That is the only reason to convert anything before sending it.

Where this tool stops

It gives you the words. It does not label who is speaking, cut clips or burn captions onto video — those are Slice’s paid tools. If you need subtitle files specifically, the same transcriber is set up for that on audio to SRT, and if you already have subtitles in the wrong format, the subtitle converter does that in your browser without uploading anything.

More Free Tools

Shorts by Slice

From the team behind these free tools

One video. A week of posts.

Upload an episode and get back the moments worth posting, cut and captioned.

  • Reads the whole episode, then picks the excerpts worth posting
  • Captions burned in and framed vertical, ready to post
  • Seven days free, no card
Try Shorts by Slice