Skip to main content

Answer · How-to guides

How do I transcribe audio to text?

Upload your recording (MP3, M4A, WAV, MP4 and more), optionally choose the language and add names and terms, then wait about a minute. Read through, rename the speakers, fix any wrong words by clicking a line to hear it, and export as TXT, Word, PDF, subtitles or NVivo format. Budget about the length of the audio for checking.

Try it on your own file

Free for 3 files a day, up to 30 minutes each. Speaker labels, timestamps, and exports for Word, SRT and NVivo.

Drop audio or video here

MP3, MP4, M4A, WAV, MOV, WEBM, OGG · up to 5 hours

Before you record

Most transcript problems start at the recording. Put the microphone close to the people speaking, choose a quiet room without echo, and if you are interviewing on a call, use a tool that saves the recording rather than recording the speaker of your laptop.

Write down names, places and jargon you expect to hear. You will need them for the names and terms box.

How long does it take?

Transcription itself is fast: in our tests on 2026-11-22 a 20-minute interview took about 23 seconds and a 55-minute meeting about 69 seconds after upload.

Checking takes longer. On clear audio, reading through and fixing a few words takes less time than the recording itself. On noisy audio with crosstalk, plan on listening to the hard parts at slower speed.

Doing it by hand

If you transcribe manually, use a player that slows audio and rewinds a few seconds with one key, and type in short bursts. A common rule of thumb is about four hours of typing per hour of clear audio. Starting from an automatic draft and correcting it is much faster.

The details

Most formats work

MP3, MP4, M4A, WAV, MOV, WEBM and OGG, plus AAC, FLAC, Opus and MKV, up to 5 hours per file.

You can record on your phone

Open the site on your phone and record in the browser. No app to install.

Clean or verbatim

Clean removes um, uh and short stutters for easy reading. Verbatim keeps every word.

Speakers are labelled automatically

Rename a speaker once and every line changes.

Exports for every use

TXT for free; Word and PDF for reading, SRT and VTT for video, NVivo and MAXQDA for research on paid plans.

Step by step

  1. Get the recording file

    Find the file from your phone, recorder, Zoom or video editor. If it's a video, upload the video; it is shrunk to audio in your browser.

  2. Add language and names

    Language is detected automatically, but choose it for short or mixed-language clips. Type names and jargon in the names and terms box.

  3. Upload

    Drop the file on the page or record directly. The first part of the text shows in seconds; for files of 10 minutes or more you can close the page and we email you.

  4. Rename the speakers

    Click Speaker 1 and type a name. Every line updates. Move or merge lines where people overlapped.

  5. Fix errors by listening

    Search for names to check spelling. Click any line to play from there; slow to 0.75x for fast speech.

  6. Export

    Copy the text or export TXT, DOCX, PDF, SRT, VTT, or NVivo and MAXQDA format. Then delete the file if you don't need it online.

Common mistakes

  • Uploading a heavily noise-reduced copy instead of the original.
  • Skipping the names box when the recording has unusual names.
  • Publishing quotes without listening to them once.
  • Waiting on the page for a long file; close it and wait for the email.

Questions people also ask

How do I transcribe an audio file in Word?

Some versions of Word have a built-in transcribe feature with its own limits. Here you can upload the file, fix speakers and words, and export a DOCX with speaker names and timestamps on every paragraph to open in Word.

How do I transcribe a voice memo on iPhone?

Share the voice memo to Files, then upload the M4A file here from your phone's browser. Or record directly in the browser on this site.

Can I transcribe a YouTube video?

We can't download YouTube videos. Our YouTube transcript page fetches a video's existing captions when YouTube allows it. Otherwise, upload your own copy of the audio if you have the right to use it.

Is it free?

Yes for 3 files a day, up to 30 minutes each, 120 minutes a month, with TXT export. Paid hours start at $9.99 for 10 hours and never expire.

How accurate is automatic transcription?

On clear audio most words come out right, including names and numbers. Crosstalk, phone audio and noise cause more errors, mostly at speaker changes. Always check quotes and numbers.

Related questions