MP3 is what a voice recorder, a podcast host or a call app hands you, and it is the one audio format everything opens. Drop the file below, or paste the episode link, and take away a genuine .docx. The first transcript is free, readable end to end at any length · then packs from $8 for 1,000 minutes, with credit that never expires.
How it works
Nothing to install, no conversion first.
Drop the file into the box above, or paste a link instead: podcast hosts, YouTube and Apple Podcasts addresses with their ?i= parameter all resolve to the audio.
Every line carries a timestamp, so clicking one plays that moment back. Signed in, correct a misheard word and it is kept.
Pick DOCX from the export menu, timestamps in or out. TXT, SRT, VTT, JSON and Markdown come from the same job at no extra charge.
About MP3
Lossy, small and universal, which mostly works in your favour.
Handheld recorders, dictation apps, podcast downloads and exported call recordings default to MP3, because every device made in thirty years plays one.
Voice at 64 kbps mono runs about 29 MB an hour, stereo at 128 kbps about 58 MB. Ten hours stays far under 5 GB.
A 32 kbps phone-line recording sounds thin, but the words survive. What you lose shows up as the odd wrong name.
Speaker 1, Speaker 2 and so on sit beside the lines in the transcript view on this page. The Word download is the spoken text itself.
| MP3 bitrate | One hour | Three hours | Ten hours |
|---|---|---|---|
| 64 kbps mono (voice) | ~29 MB | ~86 MB | ~288 MB |
| 128 kbps stereo | ~58 MB | ~173 MB | ~576 MB |
| 320 kbps stereo | ~144 MB | ~432 MB | ~1.4 GB |
Approximate. Over 500 MB, start from the upload page; the ceiling is 10 hours or 5 GB.
Who needs the document
Each of these finishes in writing, not in a media player.
Draft show notes or a newsletter from the episode you published, starting from prose you can cut.
The recorder gives you the interview; the Word file gives you quotes to paste into a piece.
Turn the MP3 your call app exported into a written record of what was agreed.
Talk a report through on the way home, then edit the document it becomes.
Pricing
The first transcript costs nothing and you can read all of it, at any length. Then: $8 for 1,000 minutes, $12 for 2,000, $24 for 5,000. No subscription, and credits never expire.
A 38-minute episode costs $0.30 on the $8 pack. Twenty of them is 760 minutes, still inside that one pack. Details on the pricing page.
FAQ
Drop the MP3 into the box here, wait for the transcript, then choose DOCX from the export menu. Preview needs no signup.
Yes. A genuine .docx, not rich text renamed, so Word, Google Docs, Pages and LibreOffice all edit it normally.
No. Speaker labels appear in the transcript view on the page, beside each line. The DOCX, like TXT, SRT and VTT, carries the words only; JSON is the export that includes speaker information.
The ceiling is 10 hours or 5 GB. The box on this page takes files up to 500 MB, and anything larger continues on the upload page.
Yes. Recordings at 32 or 64 kbps, normal for phone calls and old dictation hardware, transcribe fine. Very quiet or clipped audio is the real problem.
Yes, they are optional on the DOCX and TXT downloads. Leave them off for a clean reading copy, or keep them to find a moment in the audio.
Yes. Podcast host links, YouTube links and Apple Podcasts episode links all work. Spotify, Podbean, Kick and Zoom links are refused, so upload the MP3 for those.
The same transcript exports as TXT, DOCX, SRT, VTT, JSON and Markdown, and taking one again later costs nothing.
Instant preview, no signup, first transcript free at any lengthours.