How to Transcribe WhatsApp Voice Messages to Text on iPhone (Any Language, Offline)
WhatsApp’s own voice-to-text is missing for many countries and languages and never covers saved files. Here is a two-tap way to transcribe any voice note on the iPhone itself, with nothing uploaded.
Long voice messages arrive at the worst moments: in a meeting, on a train, in a library. WhatsApp added transcripts for some people, but the feature depends on your country and language, and it does nothing for voice notes you have already saved or received in another app. This guide shows how to transcribe any voice message to text on an iPhone, in any major language, without uploading the audio anywhere.
Short version: long-press the voice note, tap Share, choose Copy to Voice to Text, wait a few seconds, read. Voice Note to Text runs the speech model on the phone, detects the language automatically, and costs $2.99 once, with no subscription.
Why the built-in transcript is often missing
The native feature is rolled out by region and by language, and it only works inside the chat on the device that received the message. If your language is not on the list, if the message was forwarded as a file, or if you are looking at a voice note you saved months ago, there is no transcribe button. Third-party transcription is the way around all three.
The workflow, step by step
- In the chat, long-press the voice message and tap Share.
- In the share sheet, choose Copy to Voice to Text. The app opens and starts immediately.
- The first transcription loads the bundled speech model; later ones start instantly. A one-minute note takes a few seconds on a recent iPhone.
- Read the transcript on screen, then tap Copy to paste it into a reply or a note, or Share .txt to save it as a text file.
You can also save voice notes to Files first and pick them from inside the app. The app opens .opus directly, so there is no need to convert WhatsApp audio to MP3 before transcribing it. .m4a (iPhone Voice Memos, voicemail), .mp3, .wav, .ogg (Telegram, Signal) and .caf (iMessage audio) all work too.
Language detection
There is no language setting. The model listens to the first seconds of audio and decides. Around a hundred languages are covered, including English, Spanish, German, French, Portuguese, Italian, Dutch, Turkish, Arabic, Hindi, Tamil, Bengali, Indonesian, Japanese, Korean and Chinese. Mixed-language messages are transcribed in whichever language dominates.
Privacy: how to check that nothing is uploaded
Most transcription apps send audio to a cloud service. That is a problem for a private voice note. On-device transcription is easy to verify: switch on airplane mode, share a voice note to the app, and watch it transcribe anyway. The speech model ships inside the app bundle, which is why the download is around 60 MB and why there is no account or subscription.
Getting better results
- Voice notes recorded close to the microphone transcribe best. Wind, traffic and echo lower accuracy, as with any speech recogniser.
- For a very long recording, expect proportionally longer processing. Ten minutes is usually done in well under a minute.
- If you need the audio itself in a standard format, convert it with Voice Note Audio Converter instead.
- To translate a transcript, copy it into Apple Translate or any translator. Transcription and translation are separate steps.
The apps mentioned here are independent utilities built by the author. They are not affiliated with, endorsed by, or connected to WhatsApp LLC or Meta Platforms, Inc. WhatsApp is a registered trademark of Meta Platforms, Inc.