How to Transcribe Voice Memos and Voice Messages (iPhone, Android, Desktop)

Voice memos and voice messages are the fastest way to capture an idea — and the slowest way to review one. Reading a transcript takes seconds; listening again can take minutes. This guide shows how to get the audio file out of your phone or chat app and turn it into text for free, without sending a private recording to a server.

The short version

1. Save the voice memo or voice message as a file (steps for each app below).
2. Open Voice note to text on your phone or computer.
3. Drop the file in, choose the spoken language, and wait a few seconds.
4. Copy the text, or download it as TXT or SRT.

The transcription runs inside your browser using the open-source Whisper speech model. The first run downloads the model (about 40–250 MB depending on the size you pick); after that it is cached and starts almost instantly.

Step 1: Get the audio file

iPhone Voice Memos

1. Open Voice Memos and tap the recording.
2. Tap the ••• button, then Share.
3. Choose Save to Files (or AirDrop it to your Mac).

The file is saved as .m4a. On a Mac, you can also open the Voice Memos app and drag a recording straight to the desktop.

Android recorder apps

Most recorder apps have a Share option on each recording. Share it to Files or Drive, or open the recordings folder in your file manager. Files are usually .m4a, .mp3 or .wav.

Voice messages in chat apps

In WhatsApp, Telegram, Signal and similar apps, long-press the voice message and choose Share or Forward, then save it to Files (iPhone) or Downloads (Android). In desktop apps, right-click the message and choose Save as or Download. These files are usually .opus or .ogg — no conversion is needed.

Step 2: Transcribe it

Open the voice note to text tool and drop the file onto the upload area (or tap it to pick a file on mobile). Before you start:

• Pick the spoken language instead of auto-detect if you know it. Short clips give the detector very little to work with, so choosing the language avoids mistakes.
• Choose a model size. Fast is fine for clear speech in English. Balanced is a good default. Accurate handles accents, noise and less common languages better but downloads more and runs slower, especially on phones.

Text appears as it is recognised. When it is done, use Copy to paste it into notes or a message, or download a TXT file.

Tips for better results

• Record close to the microphone and away from fans, traffic or music.
• If a long memo comes out messy, try the Accurate model before re-recording.
• For many memos, a desktop browser (Chrome or Edge) is much faster because it can use the graphics card.
• Names and jargon are the most common errors — a quick read-through fixes them.

Doesn’t my phone already do this?

Some phones and apps now show transcripts for certain languages. They are convenient when available, but coverage depends on your device, app version and language, and you usually cannot export the result with timestamps. A browser tool works the same on any device and any language Whisper supports, and lets you save TXT or SRT files.

Is it private?

Yes. The audio is decoded and transcribed on your own device; it is never uploaded. Only the model files are downloaded from the internet. You can read more in our privacy policy.