Short answer: Yes, Apple transcribes voice memos. Since iOS 18, recordings you make in Voice Memos, Notes, and (with Apple Intelligence) recorded phone calls get an on-device text transcript, for free, with nothing uploaded. It is genuinely private and well integrated. The gaps show up at the edges: it only runs on Apple hardware, it handles about ten languages, it is built around recordings made inside Apple’s own apps, and it exports no subtitle files.
Most people asking “does Apple transcribe voice memos” have a recording already sitting on their phone and want the words out of it. For that exact case, Apple has you covered. The friction starts when your audio came from somewhere else, or when someone on your team is on Windows.
Voice Memos transcription (iOS 18 and later, on-device)
This is the headline feature. Apple added transcription to Voice Memos in iOS 18 (2024). Open any recording in the app and a transcript appears right below the waveform. You can read it, scroll it, and copy the text.
Two things worth knowing:
- It works on iPhone 12 or later, and it does not require Apple Intelligence. So a fairly wide range of iPhones can do this today.
- It runs entirely on-device. Your audio never leaves the phone for transcription. For sensitive recordings, that is a real privacy advantage over any cloud service.
On the Mac, the same feature exists but with tighter requirements: you need macOS 15 Sequoia or later and an Apple Silicon Mac. Intel Macs are left out entirely. If you are on an older MacBook, Voice Memos transcription is simply not available.
Notes and Phone call transcription
Voice Memos is not the only Apple app that transcribes.
Notes (iOS 18): You can record audio directly inside a note and get a transcript, again on iPhone 12 or later. Notes is also more flexible than Voice Memos about imported files: you can attach an audio file to a note using the paperclip, and Notes can transcribe some of those attachments. If you are trying to get text out of a file Apple did not record, Notes is often the more forgiving path.
Phone and FaceTime call recording (iOS 18.1, October 2024): When you record a call, it lands in the Notes app with a transcript, and on Apple Intelligence devices you also get an AI summary. This is on-device and end-to-end encrypted. Region and language coverage vary, and the other party is notified that recording has started.
Newer still, iOS 26 and macOS 26 Tahoe introduced a new SpeechAnalyzer and SpeechTranscriber framework aimed at long-form audio. Independent tests found it faster than Whisper, roughly 55% faster on speed alone. That is a speed comparison, not an accuracy or language-breadth one, and it does not change the fundamentals below: it is still on-device, and it still exports no SRT or VTT.
The arbitrary-file gap (the key limitation)
Here is where a lot of people get stuck. Apple’s transcription is built around recordings made inside its apps. Importing an arbitrary file is where the path gets narrow.
Voice Memos will not import an MP3 or an MP4 directly. It accepts M4A (AAC) only. So if a colleague sends you an MP3 interview, or you download a recording off the web as an MP4, Voice Memos rejects it as-is. The workaround: convert the file to M4A first (QuickTime can do this), then drag the M4A into Voice Memos. It works, but it is a manual step every single time.
Notes is more permissive and can attach and transcribe some audio files without that conversion dance, so it is not accurate to say “Apple can only transcribe its own recordings.” The honest framing is this: Apple’s ecosystem can transcribe some imported files, but the path is app-dependent, format-restricted (Voice Memos is effectively M4A-only), and clunky compared to a tool where you drag any MP3, MP4, or WAV in and get text back.
If your workflow is “I record on my iPhone and read the transcript,” none of this matters. If your workflow is “I have a folder of mixed audio files to transcribe,” it matters a lot.
Apple-only, and nothing on Windows
Every feature above lives inside Apple’s operating systems. There is no Windows version of any of it. No Voice Memos transcription on a PC, no Notes transcription, nothing.
And even within the Apple world, the Mac side needs Apple Silicon. So the full requirement stack is: an iPhone 12+ on iOS 18+, or an Apple Silicon Mac on macOS 15+. If your team is mixed Mac and Windows, or you are on an Intel Mac, Apple’s built-in transcription is not a shared solution.
Languages and the missing subtitle export
Two more practical ceilings.
Languages. Voice Memos transcription covers roughly ten languages: English (all variants), Spanish, Portuguese, Italian, French, German, Japanese, Korean, Simplified Chinese, and Traditional Chinese. Not every language is available in every region. That list handles a lot of the world, but it is a fraction of what an open speech model reaches.
Subtitles. Apple transcription outputs plain text with no word-level timings. There is no SRT, no VTT, no way to line up text with the moment it was spoken. As of iOS 26, Voice Memos export is still copy-paste only: no TXT, SRT, or DOCX file export. For a personal note that is fine. For captioning a video, it is a dealbreaker, because subtitles are defined by their timings.
Apple vs SnailText
For the common case (a recording you made on your iPhone), Apple is the simplest tool that exists and you should just use it. The comparison below is for the cases Apple does not cover.
| Apple (Voice Memos / Notes) | SnailText | |
|---|---|---|
| Platforms | Apple only; Mac needs Apple Silicon | Mac and Windows |
| Runs on-device | Yes | Yes |
| In-app recordings | Yes, core strength | Yes |
| Import any MP3 / MP4 / WAV | Restricted (Voice Memos is M4A-only) | Yes, drag any file |
| Folders / batch | No | Yes, folders and drag-drop |
| Languages | ~10 | ~99 (Whisper) |
| Subtitle export (SRT / VTT) | None | SRT, VTT, plus TXT, MD, JSON |
| Cost | Free, built in | Free tier (unlimited compact models) |
| Speaker labels (diarization) | No | No |
SnailText is a local file-transcription app for both Mac and Windows. It runs Whisper (and Parakeet TDT on Pro) on your own machine, so audio stays local the same way Apple’s does. You drag in any MP3, MP4, WAV, or downloaded recording, pick a language from about 99, and export SRT, VTT, TXT, MD, or JSON. The free tier is unlimited on the compact models with no account. Pro (Whisper large and Parakeet) is $7.49/month or $89/year. It does not do speaker diarization, so if you need “who said what” labels, it is not the tool for that yet.
If you want to try the file workflow, drag a recording into SnailText and see the text and subtitle export before deciding.
The honest verdict
Does Apple transcribe voice memos? Yes, and it does it well for what it is built for: recordings you make on an iPhone or Apple Silicon Mac, transcribed privately on-device, at no cost, with zero setup. If that describes your situation, stop here and use Apple.
Reach for something else when you hit one of Apple’s four walls: you are on Windows or an Intel Mac, you need to transcribe imported MP3 or MP4 files without converting each one, you work in a language outside Apple’s ten, or you need SRT/VTT subtitles with timings. Those are exactly the gaps a cross-platform local tool like SnailText is meant to fill, without giving up the on-device privacy that makes Apple’s version worth using in the first place.
Feature availability reflects iOS 18 through iOS 26 and macOS 15 through 26 as of mid-2026. Apple updates these tools regularly, so check current OS release notes for the latest language and region coverage.
Related reading:
- Transcribe audio and video files locally - drag any file in, export SRT/VTT/TXT
- Free voice to text - if budget is the main constraint