Why WhatsApp's own transcript feature is phone-only
WhatsApp does have voice message transcripts. Its help article, "How to turn voice message transcripts on or off," offers platform tabs for Android, iOS, Web, Windows and Mac, and the difference between them is the whole story: the Android and iPhone tabs carry real, numbered instructions, while the Web tab says only that voice message transcripts are "only available on certain devices and may not be available to you yet," and recommends keeping WhatsApp updated so you get the feature when it arrives.
The reason is architectural rather than arbitrary. WhatsApp states that transcripts are generated on your device, and that your personal voice messages stay protected by end-to-end encryption — nobody outside the chat, not even WhatsApp, can listen to your voice messages or read the transcript. On a phone, that means leaning on the operating system's own speech recognition. On iPhone, the help page says you need iOS 16 or newer with Siri enabled, and that you may have to download the language pack when you turn the feature on. A browser tab has nothing equivalent to lean on.
Two more details worth knowing, because they explain a lot of confusion. Transcripts are off by default and have to be switched on per device. And only the person receiving a voice message can see a transcript of it — the sender cannot.
Supported languages, according to WhatsApp
The language list is not the same across phones, which is why two people can compare notes and reach opposite conclusions:
| Platform | Languages WhatsApp lists for transcripts |
|---|
| Android | English, Portuguese, Spanish and Russian |
| iPhone, iOS 16+ | English, Arabic, Chinese, French, German, Italian, Japanese, Korean, Portuguese, Russian, Spanish and Turkish |
| iPhone, iOS 17+ | The iOS 16 list plus Danish, Dutch, Finnish, Hebrew, Malay, Norwegian, Swedish and Thai |
| WhatsApp Web | No transcript setting; the help page says the feature is only available on certain devices |
WhatsApp also publishes a page for when a transcript refuses to appear at all. The reasons it lists are that your transcript language setting does not match the language of the voice message, that some words were not recognized (background noise is the example it gives), or that the language of the voice message is not supported.
The manual route on a computer
If you want text without installing anything, the mechanics are straightforward. WhatsApp Web downloads received voice messages automatically, so the audio is already there; you save it out of the chat from the message menu, then upload the file to whatever transcription tool you use, wait, and paste the result wherever you needed it.
It works, and for one important recording it is perfectly reasonable. The friction shows up at volume. Every audio becomes a four-step errand, the text ends up in a different window from the conversation it belongs to, and you are handing a private recording to a third-party service you may not have vetted. If it is a client's voice note, that last point deserves a minute of thought before you drag the file anywhere.
Two native controls help in the meantime: clicking the 1x badge on a playing voice message steps the speed to 1.5x and then 2x (you cannot go below 1x), and dragging the dot along the waveform jumps to a specific moment instead of replaying from the start.
What a browser extension does differently
An extension for WhatsApp Web changes the shape of the task rather than the tool. Instead of exporting audio to somewhere else, it adds a transcribe control to the voice note in the chat, and the text renders right there, attached to the message, with the sender and the timestamp still around it. No download, no upload dialog, no second window, nothing to paste back.
That matters most in the cases people actually complain about: a voice note that lands during a meeting, a five-minute ramble containing one address, a group where four people answered by audio while you were away.
Accuracy and languages
Our own extension, Kaption AI for WhatsApp Web, transcribes voice notes in the language they were recorded in — English, Spanish, Portuguese, French, German, Italian, Turkish, Chinese, Japanese and dozens more — and live translation covers more than 90 languages if the message arrives in a language you do not read.
We would rather be straight about the ceiling: as our terms say, AI transcription accuracy may vary. Speech recognition, WhatsApp's own included, is affected by background noise, crosstalk, heavy accents and people who record while walking down a busy street. WhatsApp says as much on its own troubleshooting page. Expect a very usable transcript in ordinary conditions and expect to play the audio when a specific number or a name really matters.
Privacy: what leaves your computer
This is the part worth reading closely for any extension, ours or anyone else's.
Kaption runs only on web.whatsapp.com and works with the chats you already have open there. Your WhatsApp history is kept in encrypted storage on your own device, inside the browser, so search and transcripts work without uploading your conversations — no message content is stored on Kaption's servers. Audio you choose to transcribe is sent for processing and discarded once the text comes back, and your content is never used to train AI models, never sold and never licensed.
Transcription runs first on inference servers we operate. When our own capacity is unavailable, and for translation, summaries and chat analysis, the request goes to one of a small, fixed set of AI infrastructure providers acting as our processors: Google, Groq, DeepInfra and Cerebras. What reaches them is only the audio or text you chose to process plus a language hint — no name, no phone number, no contact list, no account identifier. Data at rest is encrypted with AES-256 and everything in transit goes over TLS.
Step by step on WhatsApp Web
- Install the extension. Open the Kaption AI listing in the Chrome Web Store and click "Add to Chrome," then confirm "Add extension." On Microsoft Edge, use the Edge Add-ons listing instead; it works on Chrome, Edge and other Chromium-based browsers.
- Open web.whatsapp.com and reload it. Kaption appears inside your chats. Pin it from the puzzle-piece icon in the toolbar so it stays one click away.
- Find a voice note and transcribe it. The transcribe control sits on the voice message itself. The text appears under the message in the chat.
- Do something with the text. Play the audio and follow along, copy the transcript, summarize it, or translate it.
Installing is free and needs no credit card. The optional Basic and Pro plans add unlimited transcriptions, unlimited translation sessions, smart transcription and more scheduled messages. About 50,000 people use the extension today, with a 4.8 rating across roughly 4,700 reviews on the Chrome Web Store.
Tips for long or multi-speaker audio
Summarize first, read second. For anything over a couple of minutes, a summary tells you whether the full text is worth your attention. We wrote a whole walkthrough of that habit in how to summarize a 10-minute voice note in 3 seconds.
Treat group audio as a queue, not a stream. Consecutive voice messages auto-play in order in WhatsApp Web, which is how twelve minutes disappear. Transcribing them one by one and skimming is usually faster than listening at 2x.
Expect names and numbers to be the weak spot. Product codes, street numbers and unusual proper nouns are where any speech engine slips. When a detail is load-bearing, scrub back on the waveform and confirm by ear.
Ask for a resend when the recording is genuinely bad. If someone recorded next to a blender, no tool will rescue it — WhatsApp lists background noise as a reason its own transcripts fail. Asking for a ten-second re-record is faster for both of you.
If you would like the wider picture, including the phone-side options, we cover them in how to convert WhatsApp audio to text messages.
FAQ
Can WhatsApp Web transcribe voice messages on its own?
Not today. WhatsApp's help article about voice message transcripts has a Web tab, but it only says the feature is available on certain devices and may not be available to you yet, with no setting to switch on. The step-by-step instructions live under the Android and iPhone tabs. On a computer you need either a separate transcription tool or a browser extension that adds the capability to WhatsApp Web.
How do I turn on WhatsApp voice message transcripts on my phone?
On Android, open WhatsApp Settings, tap Chats, turn on Voice Message Transcripts, tap Choose language and pick the language most of your voice messages are in, then tap "Set up now" or "Wait for Wi-Fi." On iPhone the path is the same through Settings and Chats, and the feature needs iOS 16 or newer with Siri enabled. To read a transcript, tap and hold the voice message and choose Transcribe.
Which languages can WhatsApp transcribe?
It depends on the device. WhatsApp's help page lists English, Portuguese, Spanish and Russian for Android. For iPhone it depends on the iOS version: iOS 16 and newer covers twelve languages including Turkish, and iOS 17 and newer adds eight more. If your language is not on your platform's list, a transcript will not appear at all.
Is transcribing WhatsApp audio on a PC free?
Yes, if you use the free tier of a tool built for it. Installing the Kaption extension is free and you can start transcribing voice notes immediately without a credit card; the optional Basic and Pro plans exist for heavy use. Downloading the audio and using a separate transcription service can also be free, depending on that service's limits.
Can the sender tell that I transcribed their voice note?
WhatsApp's own help page is clear that the sender cannot see transcripts — only the recipient can. With a browser extension, the transcription is something you trigger in your own tab. What the sender does see is unchanged: read receipts for voice messages follow whatever your WhatsApp privacy settings already say.
How accurate is it, really?
Good enough to skim and act on, not good enough to quote a phone number without checking. Accuracy varies with the recording: clean audio from one speaker transcribes almost perfectly, while overlapping voices, heavy background noise or an unusual accent will produce gaps. WhatsApp names the same causes on its own troubleshooting page for missing transcripts.
