Someone sends you a nine-minute voice message. You are on a train, in a meeting, or simply not in the mood to listen to nine minutes of anything. This guide shows how to move a voice message out of a chat app and into a transcript on your iPhone: one share, a short wait, and then text you can skim in thirty seconds, search next month, and quote back accurately.
The message is transcribed on your phone and it works offline, which matters when the message is personal. There is no account, and the audio is never sent to us.
Why long voice notes are the actual problem
Short voice messages are fine. Nobody needs a transcript of "on my way, five minutes". The pain is specific and it is about length:
- You cannot skim audio. A page of text can be read in any order at any speed. A voice message plays start to finish, and the one sentence you need is somewhere in the middle.
- It runs at the sender's pace. Playback speed controls help a little, and then the sender starts thinking out loud and you are back to real time.
- You cannot search it. Three months later you remember that someone told you a price, a date or an address in a voice note. Scrolling a chat looking for a grey audio bubble is not a search.
- You cannot quote it. Repeating a number back correctly means replaying the same eight seconds four times.
- It needs your ears free. Open-plan office, quiet carriage, sleeping baby, no headphones. Text does not.
- It is harder in a second language. Reading a language you are still learning is far easier than parsing it through a phone speaker and a bad connection.
Add to that the practical case people rarely mention: voice messages are often where the actual agreement lives. A contractor describes the work, a client changes the brief, a family member gives instructions about an appointment. That belongs somewhere more durable than a chat thread you will scroll past, which is the case for voice message transcription on iPhone rather than another replay.
How to transcribe a voice message on iPhone
- In the chat app, touch and hold the voice message and open the iOS share sheet (through Share, or Forward and then the Share button).
- Choose ThinkScribe. The first time, it may not be in the row of icons: swipe to the end, tap More, and turn it on. ThinkScribe confirms "Added to ThinkScribe. Open the app to transcribe."
- Open ThinkScribe. If the message is not in your default language, check the language chip on the Record tab first (28 languages; use Auto-detect when it is offered).
- When the transcript is ready, open it from the Library and read it on the Transcript tab.
- Tap Summary for the short version with its action items, or Chat to ask a question about the message.
- To keep it, tap ... > Export (PDF, Markdown, TXT and more), or Copy Text to paste into a reply.
Voice notes from WhatsApp and Telegram work, as do audio attachments from other apps. The app-by-app details are below.
From a WhatsApp voice message
Touch and hold the voice message, then choose Forward and tap the Share button at the bottom of the screen, or use the Share option directly if your version has one in the menu. Pick ThinkScribe. WhatsApp voice notes usually arrive as OGG audio, which imports as is. Transcribe WhatsApp voice notes on iPhone covers WhatsApp in more depth.
From a Telegram voice message
Touch and hold the voice message and choose Share or Forward, then Share, and pick ThinkScribe from the sheet. Telegram voice messages are also typically OGG. A file someone sent as an audio attachment rather than as a voice message shares the same way.
From a Signal voice message
Touch and hold the message and use the share or save option, then pick ThinkScribe from the sheet, or save the file to Files first and import it with Import Audio File on the Record tab. Signal voice messages generally arrive as M4A audio.
From Messages, Mail, Voice Memos or Files
An audio message in Messages, a recording attached to an email, a memo in Voice Memos, or anything sitting in Files: tap the Share button and choose ThinkScribe. Voice Memos recordings are M4A and import directly.
Speech recognition runs on the iPhone, which is why the whole thing also works in airplane mode, in a basement, or abroad with data switched off. The offline speech-to-text explainer covers how that works on a phone.
Read it instead of listening to it
The transcript lands in the Library, and this is where the nine minutes turn into thirty seconds:
- Skim it. Read the whole thing at your own pace and stop at the part that matters. The transcript shows cleaned text by default, with hesitations and fillers removed, which shortens a rambling message considerably; Show Original brings back every word.
- Tap a line to hear it. If a name or a number looks wrong, tap that line and hear exactly that moment rather than hunting through a waveform.
- Search the library. The Search transcripts field searches the full text of every transcript, so the price from three months ago is one search away instead of a scroll through a chat. Search inside audio recordings shows how far that goes.
- Summarize a long one. Choose Summarize for a summary with an Action Items list. For a fifteen-minute brief from a client, that is often all you need before replying.
- Ask about it. Ask questions like "What decisions were made?" about that one message.
- Keep it properly. Export to PDF for a project folder, Markdown for your notes, or plain text to paste into a reply. The export guide lists every format.
Voice notes with more than one voice
Some voice messages are not one person talking: someone passes the phone around, records a room, or forwards a snippet of a conversation. Speaker separation runs on imported files exactly as it does on your own recordings, so the transcript arrives split by voice. Name a speaker once and the same voice is recognized next time. The speaker diarization article explains what that can and cannot do, particularly when two people talk over each other.
Why on-device matters for this specific thing
Voice messages are, on average, the most personal audio on your phone. They are people's health, their money, their arguments and their plans, recorded in their own voice, sent to you and not to anyone else. Uploading them to a transcription service to save yourself nine minutes is a real decision, not a formality, and the sender is not there to be asked.
Transcribing on the phone removes most of that decision. ThinkScribe needs no account, and your recordings and transcripts are never sent to us, so there is no copy on a transcription service and no retention policy to read. The on-device versus cloud comparison makes the general case; for voice messages from friends and family it is close to the whole argument.
When the chat app's own transcript is enough
Some messaging apps now transcribe voice messages themselves, for a list of languages you download. For a short message in a supported language, that is the fastest route and there is nothing to install. Its limits are scope: one message at a time, copy as the only export, no timestamps, no speaker names, no searchable library, and nothing for audio that arrived as a file rather than as a voice message. Three ways to transcribe voice messages on iPhone walks through those built-in options before the share route, and Voice Memos transcription sets out the same boundary for Apple's own recorder.
A reasonable rule: use the built-in transcript for anything under a minute, and share anything longer, anything you might need again, or anything in a language the built-in option does not cover.
Frequently asked questions
How do I turn a voice message into text on iPhone?
Touch and hold the message in the chat app, open the iOS share sheet through Share or Forward, and choose ThinkScribe. Open the app and the audio is transcribed on the phone; then you read, search, summarize or export the text. Some chat apps also have their own built-in transcripts for short messages in supported languages.
Does transcribing a voice message send it to a server?
Not with ThinkScribe. Speech recognition runs on the iPhone itself, so transcription works with no connection, and the audio and transcript are never sent to us. With a cloud transcription service the audio is uploaded and processed on someone else's computer, which is worth knowing when the message was not yours to share.
Can I transcribe a very long voice message?
Yes. A forty-minute voice note is fine; it just takes longer than a two-minute one. For anything that long, the summary and its action items are usually the fastest way in, with the full transcript underneath when you need the exact words.
What audio formats do voice messages use?
It depends on the app. Voice notes from WhatsApp and Telegram are usually OGG, Signal voice messages are generally M4A, Voice Memos saves M4A, and audio attachments can be almost anything. MP3, M4A, AAC, WAV, AIFF, FLAC, CAF, OGG and Opus all import, as do AMR, WMA and most other formats, so in practice you can share whatever arrives without converting it first.
