Turn a WhatsApp voice message into text

Someone sends four minutes of voice note and you are somewhere you cannot listen. Turning it into text takes one share and a few seconds — and on an iPhone the audio does not have to travel anywhere to make that happen. Kaseta transcribes it on the phone, in 99 languages, with no account and nothing uploaded.

That last part is not a detail here. A voice note is someone else’s voice, sent to you in a private chat. Handing it to a web uploader means making a decision about their audio on their behalf. On-device transcription is the version of this where you do not have to.

Try WhatsApp’s own transcripts first

Recent versions of WhatsApp for iPhone transcribe voice messages themselves, on the device, in a short list of languages. It is under Settings → Chats. If it is there and it handles the language you need, that is the whole answer and this page is over — no download required. Kaseta is for the cases it leaves out: a language outside that list, a long recording, or audio that arrived by email rather than by chat.

When something else is the better tool

You are on Android or a desktop. Kaseta is an iPhone app. WhatsApp Web plus one of the browser transcription tools will serve you better, with the usual caveat that the file gets uploaded.
The voice note is a work meeting. If a team needs the transcript, with speaker labels and a shared searchable archive, a cloud service like Otter is built for exactly that and Kaseta is not.
You just need the gist of one message. Play it at double speed. Genuinely faster than installing anything.

For voice memos you record yourself rather than receive, the voice memo guide covers the same ground, and the homepage has the short version of how the app works.

Steps

Share the voice message

Press and hold the voice note in the chat. WhatsApp offers to forward or share it; sharing opens the iOS share sheet, which is how the audio reaches another app.

Choose Kaseta

Pick Kaseta from the share sheet. The file is read where it already is, on the phone — it is not sent anywhere for processing.

Read instead of listening

The transcript is produced on the device, so a four-minute voice note becomes something you can skim on a train with no signal.

Keep the part that matters

Copy the text into any app on any plan, or export it as .md or .txt with timestamps on Pro.

Frequently asked questions

Can ChatGPT transcribe a WhatsApp voice note?

Yes. If you export the audio file and hand it to ChatGPT, it will return text, and for one message that is a perfectly reasonable answer. The trade is that the voice note is uploaded to a server and processed there. Kaseta does the same job on the phone, which matters when the message was not yours to share.

Does WhatsApp transcribe voice messages by itself?

Recent versions on iPhone can, in a limited set of languages, and the transcription runs on the device. Look under Settings, then Chats, for voice message transcripts. If the option is there and it covers your language, use it — it is free and already installed. A separate app is for what it does not reach: other languages, long recordings, and audio from outside WhatsApp.

What about voice notes in another language?

The transcription engine covers a long list of languages, and a voice note is transcribed in the language it was spoken in — a message in Portuguese comes back as Portuguese text, not as mangled English. That list belongs to the engine; the app's own interface ships in far fewer languages, and the two numbers are not the same claim.