How to transcribe voice memos to text (iPhone & Android)

Want to transcribe voice memos to text without retyping them by hand? Get the audio file off your phone (thirty seconds of work on iPhone or Android), upload it to FileToText, and you'll have an accurate transcript a few minutes later. The first 10 minutes of every file are free, so you can test the transcript quality on a typical memo before paying anything.

The short version

• iPhone voice memos are M4A files. Most Android recorders also produce M4A or MP3. FileToText accepts both, plus WAV, AAC, OGG, FLAC, and common video formats.
• Create a free account (Google, Apple, or email, no credit card) and the first 10 minutes of any file are transcribed free. A memo of 10 minutes or less is free in full.
• Longer files: $3 per audio hour, billed per started hour, $3 minimum, paid once per file. Or a subscription from $10/month (600 minutes) if memos pile up.
• 90+ languages, auto-detected. Speaker labels show who said what.
• Accuracy: 95%+ on clear audio; noise, heavy accents, and overlapping speakers bring that down, and we'd rather say so upfront.
• Exports: TXT, DOCX, SRT, WebVTT, or a read-only share link.

Step zero: get the memo off your phone

The transcription part is easy. The mildly annoying part is that voice memos live inside a phone app, so first you need the memo as a file you can upload.

On iPhone

The Voice Memos app stores everything as M4A, a compressed but perfectly transcribable format (we cover it in detail in how to transcribe M4A files). Open the memo, tap the share icon, and pick a route:

• Save to Files drops it into iCloud Drive or on-device storage.
• Google Drive or Dropbox appear in the share sheet if you have their apps. FileToText imports directly from both, so this is the shortest path.
• AirDrop to a Mac, or email it to yourself for short clips.

If iCloud sync is on, memos also appear in Voice Memos on your Mac, where you can drag them straight out of the app.

On Android

Android is more varied, but the pattern is the same. As of mid-2026, Google's Recorder app on Pixel and Samsung Voice Recorder both save M4A by default. Open the recording, hit Share, send it to Google Drive or Dropbox, then import from there. A USB cable and your computer's file browser work too.

One caution: some older or third-party recorder apps default to AMR or 3GP. FileToText takes MP3, WAV, M4A, AAC, OGG, and FLAC, so check your recorder's settings and switch it to M4A or MP3 before the next recording that matters.

How to transcribe voice memos in FileToText

1. Create a free account. Google, Apple, or email. No credit card, and nothing starts billing behind your back.
2. Upload the memo. Drag the file in, or import it straight from Google Drive or Dropbox if that's where your phone sent it.
3. Add hints if the audio is rough. Language is auto-detected, but manual language and speaker-count hints help messy files: a two-person argument benefits from a "2 speakers" hint, a Dutch memo full of English jargon benefits from a language hint.
4. Wait a few minutes. A typical memo is done before your coffee is.
5. Review with the synced player. The transcript highlights word by word as the audio plays, and clicking any word jumps playback to that exact moment. Rename "Speaker 1" and "Speaker 2" to real names, turn on filler-word cleanup, and use the per-word confidence view to spot passages worth double-checking.
6. Export. TXT or DOCX for notes and records, SRT or WebVTT if the audio belongs under a video, or a read-only link if someone else just needs to read it.

What voice memos do to accuracy

Turning a voice recording to text accurately depends less on the tool than on how the recording was made. A voice memo is captured on whatever mic you had, wherever you happened to be, and that shows up in the transcript in predictable ways:

• Pocket and handling noise. Fabric rustle and hand taps can bury whole words. The words next to the noise usually survive; the ones under it don't.
• Distance. A phone on a lecture hall desk picks up the room, not the lecturer. Expect the echoey, far-away stretches to be the weakest parts of the transcript.
• Overlapping speakers. Interviews mostly take turns; arguments don't. Crosstalk is the hardest thing for any transcription engine, ours included.
• Wind and outdoors. The classic idea-memo-on-a-walk. Wind across the mic reads as a roar and can flatten everything under it.

Our honest baseline is 95%+ on clear audio. The problems above pull that number down, and we'd rather tell you that than pretend otherwise.

For memos you already have, two things help: give a speaker-count hint at upload, then use the confidence view and click-a-word playback to check the shaky spots against the audio and correct them in your exported DOCX. For future memos, get the phone within arm's length of whoever is talking, mic end toward the speaker, on a soft surface. A foot less distance does more for accuracy than any setting.

Test the quality free on your first 10 minutes

Every file you upload gets its first 10 minutes transcribed free, and that doesn't expire, so a short memo doubles as a full test of the transcript quality. It happens to fit voice memos perfectly, because most memos (an idea on a walk, one interview answer, a phone call you needed on the record) come in under 10 minutes. Upload, transcribe, export, and judge the quality for yourself.

When a recording runs longer, the math stays small. A 45-minute lecture costs $3, the minimum charge. A 2-hour interview costs $6, once, per file, and nothing renews. If you're transcribing every week, the Basic plan at $10/month covers 600 minutes. Not satisfied? Full refund, no questions asked. For a wider look at what free tiers actually get you elsewhere, see free transcription software: what free actually gets you.

Quick answers

How do I convert a voice memo to text on iPhone?

Share the memo out of the Voice Memos app (Save to Files, or straight to Google Drive or Dropbox), then upload it to FileToText. The file is an M4A and uploads as-is, no conversion step needed. Under 10 minutes, the whole transcript is free.

How do I turn a voice recording to text on Android?

Same idea: share the file from your recorder app to Drive or Dropbox, import it into FileToText, transcribe. If your recorder outputs an unusual format like AMR, switch its settings to M4A or MP3 first.

Are memo transcripts accurate enough to quote?

On clear, close-mic audio, expect 95%+ accuracy. Before quoting anyone, click the sentence in the transcript and listen: the synced player exists precisely so you can verify the words that matter.

Can I transcribe an argument or dispute I recorded?

Technically it's just another audio file: upload it, and speaker labels will keep track of who said what, though heavy crosstalk sections will be rough. Legally, recording laws vary by country and by US state (one-party versus all-party consent), and we're a transcription tool, not a law firm. Check your local rules before relying on a recording.

How long can a file be?

Up to 20 hours per file. The first 10 minutes are always free; beyond that it's $3 per audio hour, billed per started hour, or a subscription if this is becoming a habit.

Got a memo waiting? Upload it to FileToText and the first 10 minutes are on us.