M4A to text: transcribe voice memos and other M4A files

Upload an M4A file and GhostMinutes turns it into text with speaker labels and timestamps. M4A is the audio-only version of the MPEG-4 format, and it is what the iPhone Voice Memos app exports by default. So if you recorded a meeting, lecture or interview on an iPhone, this is probably the file you have. You don't need to convert it to MP3 first.

  • Upload the .m4a as it is
  • Speaker labels and timestamps
  • Export TXT, DOCX, PDF, SRT or VTT
Transcribe your M4A fileNo card. No signup.

Drag & drop your audio or video files

or Browse Files

Any audio or video format (MP3, MP4, MOV, M4A, WAV, FLAC, MKV…) · Up to 1 GB · Multiple files supported

Free accounts get the first 30 minutes of each upload. Paid plans from $1 a month transcribe the whole file, up to 1 GB.

Each file uses at least 15 of your minutes, even a short one.

59 seconds, two speakers, unedited

Sample result

This is the transcript exactly as GhostMinutes produced it, mistakes included. A video gives the same kind of result, because only its sound is transcribed. Click a timestamp to hear that line.

Community bike repair interview

0:59 · English · 2 speakers

Synthetic voices · unedited output
  1. Speaker 1

    Thanks for making time today to start how long have you been running the community bike repair program?

  2. Speaker 2

    About four years now.

  3. Speaker 2Check 1

    We started with two volunteers in a borrowed garage, and now we run three workshops a week.

  4. Speaker 1

    What surprised you most in the first year?

  5. Speaker 2

    Honestly, how many people came back.

  6. Speaker 2

    We expected one-off repairs, but most people wanted to learn how to fix the bike themselves.

  7. Speaker 1

    So the teaching part grew on its own.

  8. Speaker 2

    Right.

  9. Speaker 2

    By the second winter, we had a waiting list for the Thursday class, so we added a Saturday morning session.

  10. Speaker 1

    How do you fund it?

  11. Speaker 2

    Mostly small grants from the city plus a donation jar.

  12. Speaker 2

    Parts are the biggest cost, roughly $800 a month.

  13. Speaker 1

    If someone wants to start something similar in their neighborhood, where should they begin?

  14. Speaker 2

    Find one reliable space first.

  15. Speaker 2

    Tools and volunteers are easier to find than a place that lets you be messy every week.

  16. Speaker 1Check 2

    good advice thanks again thank you it was fun

What to check in this transcript

  1. 1 The speaker says "two volunteers and a borrowed garage". The transcript has "in a borrowed garage". One small word, the kind of slip to catch before you quote someone.
  2. 2 Both people speak here: Speaker 1 says "good advice, thanks again" and Speaker 2 answers "thank you, it was fun". The transcript puts both on one line under Speaker 1. In GhostMinutes you can move a line to the right speaker, which fixes a wrong label. A line that holds two voices still has to be split by hand in your export.

A 59-second two-voice interview made with synthetic voices, transcribed by GhostMinutes on 13 November 2026. The output is unedited, including its mistakes.

Get a voice memo off your iPhone

Apple's steps for saving a recording to the Files app, so you can pick it on the upload page:

  1. 1

    Open Voice Memos

    It lives in the Utilities folder if it isn't on your home screen. Tap the recording you want.

  2. 2

    Tap More, then Share

    The share sheet opens with the recording attached.

  3. 3

    Save to Files

    Scroll down, tap Save to Files, pick a folder and tap Save. The memo is saved as an .m4a file.

  4. 4

    Upload it

    Open this page in Safari on the same phone, tap the upload box and pick the memo from Files. A computer works too, after AirDrop or iCloud Drive.

Keep the default export

The share sheet has an Options button. Leave it alone. By default Voice Memos exports a single .m4a, mixes any layered recordings down and turns Spatial Audio into stereo, which is all a transcript needs.

Options also offers Editable, which saves a QuickTime Audio file so the effects and layers can be changed later. That is for editing in Voice Memos, not for transcription.

Got several memos? Save them all to Files, then upload them together on the upload page, which takes several files at once. The box on this page takes one.

Other places M4A files come from

The GhostMinutes iPhone app

It records and transcribes in the app, so there is nothing to export. It is on the App Store.

Telegram

If you share a memo into Telegram and it arrives as an audio file, @ghostminutes_bot transcribes it in the chat.

Zoom

Zoom saves an audio-only M4A with many recordings. It is smaller than the video and gives the same words. See Zoom transcription.

What a recording costs in minutes

Every upload uses at least 15 minutes of your balance, then one minute per minute of audio. Free uploads on the web stop at 30 minutes.

A 3-minute voice memoUses 15 minutesUses 15 minutes
A 20-minute callUses 20 minutesUses 20 minutes
A 50-minute lectureFirst 30 minutes transcribed, uses 30All of it, uses 50
A 2-hour interviewFirst 30 minutes transcribed, uses 30All of it, uses 120

A free account has 150 minutes a month. Lite is $1 a month for 600 minutes and Pro is $5 a month for 3,600. See pricing.

Getting a better transcript from a phone recording

Put the phone close to the person who talks the most, and away from laptop fans. The closer the microphone, the fewer words get lost.

GhostMinutes detects one language per file. If a memo switches language halfway through, split it and upload the parts separately.

Speaker labels work best when people take turns. When two people talk over each other, their words can land under one label. You can reassign a line to the right person afterwards.

M4A transcription questions

No. Upload the .m4a as it is. Converting first adds a step, and re-encoding one compressed format into another can only lose quality.

Voice Memos steps follow Apple's iPhone User Guide, checked on 13 November 2026.

Try it on your own recording

Upload an audio or video file and get a transcript with speaker labels and timestamps.

Upload a recording