Audio to Text

Upload a recording and get the words back as plain text you can read here and download as a .txt file. MP3, M4A, WAV and OGG, up to 100 MB and three hours per file.

Requires a free account. New accounts get 250 welcome credits.

Free account – 250 welcome credits – Full result

250 free welcome credits

This AI feature runs on credits. Every new account starts with 250 free credits, enough for several runs.

Private by design

Files travel over HTTPS and are deleted from the server right after processing. We keep no copy of your documents.

Accepts MP3, M4A, WAV, OGG

Upload straight from your device; the result is ready to download in seconds.

Nothing to install

Runs in any modern browser on desktop and mobile. Typical files finish in seconds.

How it works

1

Upload the recording

One MP3, M4A, WAV or OGG file per run. The limits are 100 MB of file and three hours of audio, whichever you reach first.

2

Check the price first

The length of the recording is measured and the cost in credits is shown next to your balance. Nothing is charged until you confirm.

3

Read or download the text

The transcript appears on the page and saves as a plain .txt file named after the recording you uploaded.

A recording you cannot read is a recording you will not use

Interviews, client calls, site visits, lectures, the voice memo you left yourself in the car: all of it is easy to capture and painful to go back to. Finding one sentence in a 50 minute file means scrubbing through it. Text does not have that problem. It can be searched, quoted, pasted into a report and read in a fraction of the time the audio takes to play.

This tool takes one audio file and returns the speech in it as plain text. The transcript is shown on the page so you can read and copy it straight away, and it downloads as a .txt file in UTF-8 named after your recording, so interview.mp3 becomes interview.txt.

The limits, stated plainly

  • Formats. MP3, M4A, WAV and OGG. One file per run.
  • Size. 100 MB. Uncompressed WAV reaches that in minutes, compressed audio in hours.
  • Length. Three hours of audio per file.
  • Output. Plain text only. No timestamps, no speaker names, no subtitle files.

The price is ten credits per started minute, with a floor of 100 credits so that very short clips still cover the cost of running the model. Because it is billed per started minute, a file of 12 minutes and 10 seconds counts as 13 minutes and costs 130 credits. You see that number, and your balance, before anything is charged. The credit rates for every paid tool are listed on the pricing page.

What to do with the transcript

Most people paste it straight into their notes, minutes or CRM. It is also the fastest way to make a recording searchable: one Ctrl+F through the .txt beats twenty minutes of scrubbing. If the material eventually lands in a document, the rest of the site picks it up from there. AI PDF Summarizer condenses a long file into a brief, Chat with PDF answers questions about it, and Analyze Contract reads the terms if the conversation was about an agreement. Those tools all read PDFs, so save the text into a document first.

Working the other way round, with text that is already inside a file rather than inside a recording, use PDF to Text for a normal PDF or OCR PDF for a scan. And if your recordings are MP3, which most of them are, MP3 to Text is the same tool with the MP3 specifics spelled out: bitrate, long files and what to do with a video soundtrack.

Frequently asked questions

Which audio files can I upload?

+

MP3, M4A, WAV and OGG, one file at a time, up to 100 MB and up to three hours of audio. Video files are not accepted: export or extract the audio track first, then upload that.

What does a transcription cost?

+

Ten credits for every started minute of audio, with a minimum of 100 credits per file. A three minute voice note and a ten minute call both cost 100 credits, a 25 minute recording costs 250, and a full three hour file costs 1,800. New accounts start with 250 free welcome credits.

Am I charged before I see the price?

+

No. The file is measured first and the page shows you the number of minutes, the exact credit price and your current balance. The transcription only starts, and credits only move, after you confirm that quote.

What exactly do I get back?

+

Plain text and nothing else: the transcript on the page and the same text as a UTF-8 .txt file. There are no timestamps, no speaker labels and no subtitle formats such as SRT or VTT.

Which language does the transcript come out in?

+

The one that was spoken. The tool transcribes rather than translates, so a German meeting comes back as German text. If you need another language, translate the transcript afterwards.

How good is the result on a noisy recording?

+

It depends on the recording, not on the format. Clear speech close to a microphone transcribes cleanly. Crosstalk, a phone left on a table across the room and heavy background noise all cost accuracy, and unusual names and abbreviations are worth checking by hand.

What happens to my audio afterwards?

+

The upload travels over HTTPS and is deleted from our servers within 24 hours of processing. We do not keep the audio to train anything and the transcript is yours alone.

Related tools

Simple pricing: free tools, one subscription for AI

File tools are free without limits. AI features run on a credit subscription: every new account starts with 250 free credits, and credits on your balance never expire.

See full pricing iOS app coming soon to the App Store

We use essential browser storage to keep you signed in and run our tools, plus anonymous usage analytics that sets no cookies. No ads, no cross-site tracking.