For live voice typing on a Mac, start with built-in macOS Dictation: turn it on in System Settings > Keyboard, click in a text field and speak. If you dictate heavily, a dedicated dictation app adds its own workflow, local processing and formatting. A recorded audio or video file is a different job that needs transcription software.
Three Ways to Turn Voice Into Text on Mac
Search results for Mac speech-to-text mix three different jobs. Built-in macOS Dictation types as you speak, in the app you are using. A dedicated dictation app does the same live job with its own shortcut, processing and formatting. Transcription software turns a finished audio or video file into text afterward.
The first two are live voice typing. The third is not. The first question is whether you are speaking now or already have a recording.
| Method | Best for | Extra app required | Live voice typing | Existing recordings | Typical workflow |
|---|---|---|---|---|---|
| macOS Dictation | Occasional typing, short messages, trying voice input | No, it is part of macOS | Yes | Not its purpose | Click in a text field, start Dictation, speak, stop |
| Dedicated dictation app | Daily heavy use, a dedicated workflow, local-first processing, different formatting | Yes | Yes | Varies by app | Press the app's shortcut, speak, and the text appears at your cursor |
| Transcription software | Interviews, meetings, lectures, podcasts, voice memos | Yes, or a website | No | Yes | Upload or link a file, wait, then edit and export the transcript |
Does Mac have built-in speech-to-text?
Yes. macOS includes Dictation, which Apple describes as a way to speak to enter text anywhere you can type it. You turn it on once in Apple menu > System Settings > Keyboard > Dictation. To start, click where the text should go, then press the Microphone key if your Mac has one, use your Dictation shortcut, or choose Edit > Start Dictation.
Its strengths are practical. It is already on your Mac, it needs no extra app, and you can add several languages and switch between them while you dictate. The full setup and use flow is in how to dictate on a Mac, and the start keys are in the Mac Dictation shortcut guide, so this page does not repeat them.
Its limits are the ones Apple's pages state or leave open. Dictation is not available in all languages or regions, and features vary. Whether a request is processed on your device depends on your macOS version, language and Mac. Apps can differ in how they accept dictated text, so test the ones you rely on. If it will not start, see Mac Dictation not working.
When a dedicated voice-to-text app makes sense
Built-in Dictation is enough for many people. A dedicated app makes sense when one or more of these is true. This is about the category; for named apps compared side by side, see the best dictation apps for Mac.
Apps that type as you speak, as opposed to tools that transcribe a recording, are compared by need in our guide to dedicated Mac dictation apps.
- You dictate for hours a week, so small frictions in starting, correcting and formatting add up.
- You want a dedicated activation workflow built for dictation rather than a general system setting.
- You want local-first processing, with speech recognition running on your Mac.
- You want different formatting and editing, such as filler words removed, spoken corrections or reusable snippets.
- You would rather pay once than keep another subscription.
- You want more control over vocabulary, languages and behaviour than the built-in tool offers.
Pepys Dictate: local-first dictation for Mac
Pepys Dictate is local-first AI dictation software for Mac. You press a shortcut, speak in any app, and the text appears there. Speech recognition and formatting run on your Mac by default, with no account, and it handles spoken corrections such as "Friday, no wait, Monday". It asks for Microphone and Accessibility permission.
It is built to work wherever text can be entered, subject to app-specific differences, so test the apps you rely on. Pepys Dictate downloads its speech models on first launch (about 1.3 GB) and works offline once they are downloaded.
Pepys Dictate is a one-time purchase with no recurring subscription: $9.99 once for unlimited use on one Mac, plus tax where applicable, with future Mac app updates included. A free plan covers 2,000 words a week. Early users get the lowest Mac price. The founding price will increase as Pepys Dictate matures. It is available now for Apple silicon Macs on macOS 13 or later.
What if you already have a recording?
Live dictation and transcription are different jobs. Dictation types what you say as you say it, into a field on screen. Transcription takes a recorded file, such as an interview, lecture, meeting or voice memo, and returns text afterward, often with timestamps and speaker labels.
Playing a recording into live dictation is not what live dictation is built for, so use a transcription tool. Voice Memos on Mac includes a transcription view for recordings, and Apple notes it is not available in all countries or regions. For other audio and video files, Pepys Transcribe is a separate web product from Pepys Dictate: you upload a file or paste a link, and pay for what you transcribe.
Can transcription work locally?
Yes. Transcription can run on your own Mac when the app ships a speech model that runs there. Pepys Dictate does this: drop in a voice memo, recording or video, or record a memo on its Files page, and the transcript is made on the Mac with nothing uploaded. Its speech models download once first (about 1.3 GB).
On the free plan, Pepys Dictate transcribes files up to 10 minutes and records memos up to 10 minutes. Pro transcribes files of any length and records memos up to 3 hours. Speaker labels for files and memos are in beta and also run on the Mac.
Pepys Transcribe cannot run locally. It is cloud software: you upload a file or paste a link, and the audio is processed on servers by a third-party speech-to-text provider. It suits transcription from any browser, shared links and AI summaries. The privacy trade-offs are in local vs cloud dictation.
Open-source speech models such as Whisper can also run on your own computer, if you are comfortable installing and running them yourself.
Local vs cloud speech-to-text on Mac
Where the speech is processed is a separate choice from which app you use. Local processing runs the recognition model on your Mac. Cloud processing sends audio to a provider's servers, which return text. Each has advantages and trade-offs.
- Local: audio does not need to leave the Mac for recognition, it can work without a connection once the models are on disk, and speed does not depend on your network. The trade-offs are model downloads, disk space and memory, and speed that depends on your chip.
- Cloud: large models run on the provider's hardware, so they need little disk or memory on your Mac and can improve without a download. The trade-offs are that audio travels to a server, it needs a connection, and how long audio is kept is set by the provider's policy.
- Neither is automatically safer. Local processing does not make an app secure on its own, because the app can still store, sync or log text. Cloud processing is not unsafe by default, because providers publish retention and security terms you can read.
- Built-in Dictation can be either, depending on your setup. Apple points you to Keyboard settings to see whether dictation is processed on your device and not sent to Siri servers.
Can voice-to-text work offline on a Mac?
Mac speech-to-text can work offline, but it depends on the tool and the Mac. For built-in Dictation, Apple's pages differ by version. Its macOS 13 Ventura and macOS 14 Sonoma pages say that on a Mac with Apple silicon, requests for supported languages are processed on the device and need no internet connection.
Apple's current macOS pages do not repeat that statement, and its troubleshooting page says Dictation may require an internet connection. Its Feature Availability page lists on-device Dictation languages with a note that speech models must be downloaded. To check your own setup, turn off Wi-Fi, open Notes and try Dictation.
Dedicated apps vary too. Some run models you download first and then work without a connection, and some need a connection for every request. Web transcription tools, including Pepys Transcribe, need a connection to upload your file.
Voice to text on Mac by workflow
The right option often depends on what you are writing. These are brief starting points, not rules.
- Email and documents: Apple says you can dictate text of any length, though Dictation stops after 30 seconds of silence. For a daily writing habit, a dedicated app's formatting and corrections save more editing. To draft articles or fiction by voice, see dictation for writers and authors.
- Messaging: short bursts reward a quick start more than rich formatting, so built-in Dictation is often enough. Test it in the messaging app you use.
- Coding and technical work: identifiers, symbols and jargon are where accuracy varies most. Try your own terms in each tool, and look for custom vocabulary if you dictate technical text daily.
- Notes and journalling: built-in Dictation in Notes is a low-effort start. If you talk for a long stretch and review later, a recording plus transcription may suit you better than live typing.
- Accessibility and hands-free input: Voice Control, under System Settings > Accessibility, lets you navigate, dictate and edit by voice. Apple says standard Dictation is unavailable while it is on. Shortcut-based dictation still needs a key press.
Which Mac Voice-to-Text Option Should You Use?
Pick by the job you have today, not by the product with the most features.
- Use macOS Dictation if you dictate now and then, write short text, and want nothing extra to install or pay for.
- Use a dedicated app if you dictate daily, want a dedicated workflow, local-first processing, different formatting or a one-time purchase.
- Use transcription software if you have a recording, whether audio or video, and want text with timestamps or speaker labels.
- If the dedicated path fits, Try Pepys Dictate for Mac, or see how it compares with Apple Dictation.
Last updated November 8, 2026.