MacBook voice to text: what works right now
Honest answer first: the blablaType desktop app for macOS is still in development. But you can use voice to text on Mac today with the free dictation widget on this page, no install and no sign-up.
Download for Windows- Русский
- Українська
- English
Works best in Chrome or Edge with microphone access allowed
Where blablaType for Mac stands today
We will not pretend there is a download button for macOS yet. The full blablaType desktop app, with push-to-talk dictation in any app and text typed right at your cursor, currently runs on Windows, and the macOS version is in development. We are not promising a date until it is ready. What you can do on your Mac right now: use the free speech to text widget on this page in Chrome or Edge, send audio and video files to transcription from any browser, or rely on built-in macOS dictation for short notes. Below is what each option is good for.
Free voice typing in your Mac browser
The widget on this page turns speech into text without installing anything. It uses your browser's built-in recognition, so the quality is basic, but it is free and ready in seconds on any MacBook.
File transcription already works on Mac
Transcription runs in the browser, so your Mac is fully supported today. Upload mp3, wav, m4a, ogg, mp4 or mov and get clean text back. The first minutes are free, then a small top-up balance with per-minute billing.
The desktop app, coming to macOS
On Windows it already does push-to-talk dictation in any app: hold a key, talk, and text appears at your cursor in Word, messengers or email. The same dictation software for Mac is in development.
Built for more than English
Recognition runs on our servers and is most accurate in Russian, Ukrainian and English. The app learns your corrections, takes voice editing commands, translates as you speak and starts hands-free when you say "Okay Safi".
How to dictate on your Mac right now
- Open this page in Chrome or Edge on your Mac and click the microphone button in the widget above.
- Allow microphone access when the browser asks. The free widget needs no account and no card.
- Speak naturally and watch the words appear. Keep in mind the widget runs on browser recognition, which sends audio to Google's speech service.
- Copy the finished text into any app: a document, an email, a chat. For longer recordings, use file transcription instead of dictating live.
Built-in macOS dictation vs blablaType
| Built-in macOS dictation | blablaType | |
|---|---|---|
| Accuracy beyond English | Supports many languages, but quality drops noticeably outside English | Tuned for real Russian, Ukrainian and English speech, including fast talk and names |
| Learning your words | Text replacements have to be set up by hand and nothing is learned automatically | Fix a word once and blablaType remembers it: surnames, terms, brand names |
| Editing by voice | A fixed set of system commands | Tell it to fix a phrase, shorten a paragraph or rewrite a sentence |
| Translation while you speak | Not available | Speak your language and English appears in the field |
| Hands-free start | Needs a key press to start dictation | Say "Okay Safi" and dictation starts without touching the keyboard |
Frequently asked questions
Partly, yes. The free voice typing widget on this page and file transcription both run in the browser, so they work on any Mac today. The desktop app with system-wide dictation is currently Windows only.
The macOS desktop app is in active development, and we would rather ship it solid than name a date we might miss. Until then, the free widget on this page covers quick dictation on your Mac, and the Windows version is available if you also work on a PC.
For short English notes it is perfectly fine and works system-wide. Its weak spots are accuracy outside English, no automatic learning of your corrections, limited editing commands and no translation. That is exactly the gap blablaType is built to close.
The free web widget relies on your browser's recognition, which sends audio to Google's speech service. The desktop app and file transcription send audio to our servers only for the moment of recognition, and we do not store it.