How to Transcribe a Voice Memo on Mac After Export
Hero photo by Samuel Wan on Unsplash
To transcribe a voice memo on Mac, export the recording as an M4A from Voice Memos, then run speech-to-text on that file. Apple’s app stores the audio; it does not give you a portable transcript you can drop into notes, email, or a research folder unless you copy Apple’s built-in text (where that feature exists) or convert the file yourself.
This guide covers Apple’s share and drag-to-Finder path, the built-in Voice Memos transcript on supported Macs, and how to turn an M4A into text in the browser without uploading it to Vemoir. For the wider question of going offline after models download, see offline transcription on Mac.
How to transcribe a voice memo on Mac
The reliable path is: get an M4A you control, then choose a transcription method that matches the length and whether the file may leave the device.
- Open the Voice Memos app on your Mac.
- Select the recording.
- Export a copy (Share, or drag to a Finder folder). Apple documents both in Share a recording on Mac.
- Transcribe that M4A: Apple’s in-app transcript if it is available on your Mac, the free voice memo tool for files up to 15 minutes, or import the file in the Vemoir Mac app for longer memos and speaker separation.
Do not rename or delete files inside Voice Memos’ hidden Library folder. Treat that path as a last-resort backup, not as an editing workspace.
Export the M4A from Voice Memos
Voice Memos on Mac writes AAC audio. The usual container after export is M4A, which the browser tool and the Mac app both accept (along with MP3, WAV, and, in the browser, MP4 when the audio track decodes).
Apple’s supported export methods:
- Go to Voice Memos on your Mac.
- Select the memo.
- Click the Share button and choose an app or service, or drag the recording into a Finder window to save a copy in a folder you pick (Apple Support).
For transcription, a folder on disk is cleaner: you keep the original in Voice Memos and work on the copy.
Quality lives under Voice Memos → Settings. Apple lets you choose Compressed (smaller) or Lossless (larger) (Voice Memos settings). Either format can become an M4A you drop on a tool. Lossless files are heavier; if a browser decode fails, export again as Compressed or convert to WAV.
If the memo synced from iPhone, it appears in Voice Memos on the Mac when you use the same Apple Account (see recordings on all your devices). Export from the Mac once it has finished downloading. A cloud placeholder is not a complete file.
WhatsApp voice notes are a different container. Many land as .opus or .ogg. The Vemoir browser tool refuses those. Save or share a copy as M4A or MP3 on the phone first.
M4A to text on Mac: Apple’s transcript vs a file workflow
On macOS 15 or later with Apple silicon, Voice Memos can show a transcript in the app (View a Voice Memos transcription on Mac). The feature is not available in all countries or regions. Use it when you only need to copy a few paragraphs into Notes or email.
Stay off that path when the feature is missing, the copy-out is awkward, or you need speaker labels, minutes, or a file you can rerun later. Then export the M4A and transcribe it as a file.
Apple’s in-app text is not a certified transcript. Names, numbers, and quotes need a human pass.
Turn the exported M4A into text in the browser
The free tool at /tools/transcribe-voice-memo runs a public Whisper model in the tab. The file is not uploaded to Vemoir. On first use the model downloads into the browser cache. Later runs can work with the network off once that cache is in place.
- Length15 min
Per file. The page refuses longer audio before the model runs.
- Pace3 runs a week
Successful runs. Splitting one memo into clips spends one each.
- SpeakersNot separated
Names already in the file are kept; the tool invents none.
- Live captureNo
The browser works on a file you already have.
Swiss German speech in public Whisper models usually comes out as Standard German text.
Steps:
- Confirm the file is MP3, M4A, WAV, or MP4, and 15 minutes or shorter.
- Open /tools/transcribe-voice-memo.
- Drop the file and pick the spoken language (the tool does not translate).
- Copy or download the text. Retention stays with you.
The generic transcribe recording page accepts the same limits if you already use that bookmark.
How to check that the memo was not uploaded
- Open the tool page and start a short run.
- Open Developer Tools → Network.
- Watch for a multipart upload of your audio. You should see model or asset downloads on first use, not your M4A posted to a Vemoir host.
Internet is required for that first model download. After the model is cached, you can disconnect and run again if the browser still has the weights. Closing the tab stops a run in progress.
When offline matters for a voice memo
Voice memos are often recorded on a plane or in a room with poor Wi-Fi. The browser path fits when the memo is short and you already exported the M4A. You still need the network once to fetch the Whisper weights. After that, a cached model lets you transcribe without sending the file to Vemoir. For the full picture of browser limits versus the Mac app after models install, read offline transcription on Mac.
When the memo is longer than 15 minutes
The browser page refuses a file over 15 minutes before the model runs. Splitting a long interview into clips still spends one weekly run per successful pass, and you lose a single continuous transcript.
For long Voice Memos, import the M4A in the Vemoir Mac app (capture and models). Transcription and speaker processing stay on the Mac.
If the next capture is a live call rather than a memo, record from the menu bar instead of hunting files afterward (how it works).
Want this without hunting Voice Memos folders every time? Import the next M4A on your Mac or record the call from the menu bar.
The hidden Library folder (last resort)
On current macOS versions, Voice Memos often keeps recordings under:
~/Library/Group Containers/group.com.apple.VoiceMemos.shared/Recordings/
That path is not in Apple’s user guide. Prefer Share or drag-to-Finder whenever those work. Copy a file out if Voice Memos will not export. Do not rename, move, or delete files in place.
Frequently asked questions
How do I transcribe a voice memo on Mac without uploading it?
Export an M4A with Share or drag-to-Finder (Apple Support), then use the voice memo tool in the browser. For longer memos, import the same M4A in the Mac app.
Does Voice Memos already transcribe on Mac?
On macOS 15 or later with Apple silicon, Voice Memos can show a transcript in the app, not in every country or region (Apple Support). If the feature is missing or you need a file-based workflow, export the M4A.
Where does Voice Memos store its files on Mac?
Work from the app. Apple documents sharing or dragging into the Finder. A hidden folder under Group Containers sits on many Macs. That is a backup, not a place to edit.
Can I convert M4A to text on Mac for free?
Yes, for files up to 15 minutes and up to three successful browser runs per week on /tools/transcribe-voice-memo. Apple’s built-in transcript, where available, is another free path inside Voice Memos.
Will the browser tool label speakers?
No. Speaker profiles are Mac-app features (capture).
What about a WhatsApp voice note?
Export it as M4A or MP3 first. .opus and .ogg are refused.
Does Swiss German stay Swiss German?
Usually not in the browser. Public Whisper models typically write Standard German.
What if my memo is longer than 15 minutes?
Use the Mac app import path, or stay in Voice Memos if Apple’s transcript is good enough.
Next step
Export the memo, then open /tools/transcribe-voice-memo if it is 15 minutes or shorter. For a longer file or speaker labels, import the M4A in the Mac app and keep processing on the device.
Written by
Vemoir TeamMac meeting records by Aurum Avis Labs · Oberägeri, Switzerland
We build Vemoir, a Mac app for local-first meeting records: capture from the menu bar, transcribe on your Mac, and turn calls into minutes and action items.
Related Articles
You might also be interested in these articles
Turn WhatsApp voice to text from a saved audio file
WhatsApp voice to text starts from the saved audio, not a live tap. Short files can be transcribed in the browser without upload.
Offline Transcription on Mac After Models Download
Offline transcription on Mac after the speech model is on disk: browser trial limits, local Mac apps, and what still needs the internet.
How to Import an Audio File on a Mac for Transcripts
Import an audio file on a Mac to transcribe a meeting you already recorded. Formats, the 15-minute browser path, and when the Mac app is the better fit.