How to Get a Transcript of a WhatsApp Voice Message

Long voice notes are easy to send and awkward to deal with, especially when you are somewhere you cannot play them or you need to quote what was said. This guide covers both ways to get a WhatsApp voice message as text: the transcription WhatsApp itself added, and saving the audio file so you can transcribe it with a tool that gives you an editable transcript you can export.

Start with WhatsApp's built-in transcripts

WhatsApp began rolling out voice message transcripts in late 2024. The transcription runs on your phone rather than on WhatsApp's servers, which keeps end-to-end encryption intact, and it is off until you turn it on. Open WhatsApp Settings, go to Chats, and look for Voice message transcripts. Choose a transcript language; on iPhone you may be asked to download a language pack first. After that, press and hold a voice message and choose Transcribe, and the text appears under the bubble.

Three limits are worth knowing before you rely on it.

It supports a short list of languages, and the list differs between iPhone and Android. If the sender spoke a language that is not on it, or switched between two languages, you will get a poor transcript or none at all.

Availability varies by platform, app version and region. If the setting is missing on your phone, update WhatsApp and check again, and bear in mind it may simply not have reached your region yet. The feature is built into the phone apps; if you mostly read chats on a computer, you may not find it there.

The result stays inside the chat. You cannot edit it, there are no timestamps, and getting it into a document means selecting and copying by hand, one message at a time.

For a quick look at a one-minute message in English or Spanish, the built-in transcript is the right tool and you can stop here. For anything longer, in another language, or that you need to keep, save the file.

Save the voice message as a file

Every voice note is an ordinary audio file, and WhatsApp lets you get it out of the app. The steps depend on where you are reading the chat.

On an iPhone, press and hold the voice message, tap Forward, select it, then tap the share icon in the corner instead of a contact. The iOS share sheet opens; choose Save to Files and pick a folder in iCloud Drive or On My iPhone. The same sheet can AirDrop the file to a Mac.

On Android, press and hold the message, then tap the share icon in the toolbar or open the three-dot menu and choose Share; the exact placement varies by version. Send it to Google Drive, Gmail or any other app that accepts files. Android also keeps voice notes in WhatsApp's media folders on the phone's storage, in a folder named along the lines of WhatsApp Voice Notes, which a file manager can open if you want to copy several at once.

On WhatsApp Desktop or WhatsApp Web, hover over the voice message, click the small arrow that appears to open the menu, and choose Download. This is the easiest route when the file is going to end up on a computer anyway.

The saved file is Opus audio, which WhatsApp uses because it is small and sounds fine for speech. The name usually starts with PTT or AUD followed by a date, and the extension is .opus on older files and .ogg on newer ones; both hold the same kind of audio. Some desktop players and older transcription tools do not recognise the format. If a tool refuses it, the free OGG to MP3 converter turns it into an MP3 inside your browser without uploading anything, and the MP3 will open anywhere. Skip that step for RocketScribe, which reads the Opus file directly; every conversion of an already compressed file loses a little.

Upload the file and get the transcript

Open WhatsApp voice to text and drop the file in, or upload it from your dashboard if you have an account. Choose the language or leave it on auto-detect. Auto-detect handles most single-language notes; set the language yourself when the message is short, opens with a greeting in a different language, or is in a language with close relatives, since the detector only has a few seconds of speech to go on.

A few points that matter for voice notes specifically:

• Mixed languages. Voice notes from family and colleagues often switch mid-sentence. Recordings that mix Chinese and English are handled. For Cantonese, pick it explicitly rather than relying on auto-detect, because it has its own entry separate from Mandarin.
• Long messages. A twenty-minute voice note is fine. The transcript comes back as paragraphs with timestamps, and the player underneath highlights the line being spoken, so when a word looks wrong you can click the line and hear exactly what was said.
• Several messages. Each one is a separate file and comes back as a separate transcript. If someone told you a story in eight parts, upload the eight files and export each, or paste them into one document.
• Speaker identification is rarely needed here, since a voice note is almost always one person. Leave it off unless the recording is of a conversation.

When it is done, fix any names or slang in the editor, then export as TXT to paste into a reply or DOCX for a document. The AI prompt button copies the transcript with a prompt for ChatGPT or Claude if you want a summary, and the Google Translate link opens the transcript translated when the message was in a language you do not read.

Other ways to do this

If you would rather keep the audio on your own machine, Whisper runs locally on a reasonably recent computer and reads Opus files through ffmpeg. It takes some setup and gives you plain text without an editor, which suits people who are comfortable at a command line. Dictation features in iOS and Android can also transcribe audio if you play the message out loud next to the phone, but the result is poor and it is tedious for anything beyond a sentence or two.

A note on privacy

WhatsApp's own transcription never leaves your phone. Uploading the file to RocketScribe means the audio is processed on our servers to produce the transcript, and it is kept only as long as the retention rules for your plan allow before it is deleted. If a message contains something you would not want outside the phone, use the built-in transcript or a local tool instead.

Try it on your own voice message

Save the next long voice note that lands in your chats and drop it on the WhatsApp voice to text page. It takes the .opus or .ogg file as it is and gives you editable text with timestamps.