← Back to the shelf
MENU BAR · ON-DEVICE

Transcriber

Subtitles with translation over any window, hover translation, and Russian voice dictation into English text

Download
Version
2.4
Released
6 August 2026
Size
18.3 MB
Requirements
macOS 26 (Tahoe) or newer · Mac with Apple silicon
Checksum 8a17dcdadcabbd749d01212141da3ea6469fa25e9e0b528bf10cedb41b5ce676
SCREENSHOTS

Screenshots

WHAT IT DOES

What it does

Transcriber listens to the sound of any app, or to your microphone, recognises English speech and shows it together with a Russian translation — as subtitles in a window that stays above all other windows. It also translates English text you point the mouse at: the translation and a dictionary entry appear right next to it. On top of that, it can take dictation: press a key combination, speak Russian, and get typed text — your own words, or their English translation — right where the cursor is. macOS does the recognising and the translating on your own machine, so no internet connection is needed for it.

The app has no window of its own and does not show up in the Dock — it lives as an icon in the menu bar.

One thing to know before you download it: the app translates into Russian, and its own interface is in Russian as well.

Subtitles

  • Listens to any app that makes sound — a browser, Zoom, a player — or to a microphone. Apps that are making sound right now are marked with a note in the list.
  • You keep hearing the app as usual: Transcriber listens in, it does not take the sound for itself.
  • Show the English line, the Russian one, or both at once. Switches on the fly, with buttons on the panel itself.
  • The panel stays above every window, including full-screen ones, and on every desktop. Drag it with the mouse, stretch it by the bottom-right corner; where you put it is where it will be next time.
  • Two ways to show the text: only the current phrase, or phrases piling up with the older ones scrolling upwards. Scroll back to read something and new text will not yank the view back down.
  • Font, size and colour are set separately for the English line and for the Russian one. Background opacity is a slider, all the way down to invisible.
  • A button on the panel copies the whole session's text to the clipboard.
  • The last 50 sessions stay in «Недавние транскрибации» (Recent transcriptions), where you can switch the language and copy the text.

Translating the text under the cursor

  • Hold ⌥⌘ and point the mouse at English text in any app — a quarter of a second later the Russian translation appears beside it. The keys and the delay can be changed.
  • If text is selected, the whole selection is taken — even when it covers a heading and several paragraphs. Bold and italics are kept.
  • Along with the translation you get the macOS system dictionary entry for the word under the cursor. The dictionary finds the base form by itself: point at "caching" and it gives you "cache".
  • The pin turns the panel into an ordinary window that does not disappear. That window has four blocks: the original, its translation, the definition of the selected word, and the translation of that word.
  • The original in that window can be edited — the translation is recomputed by itself.
  • Every block has its own speaker button and its own font size buttons, and the sizes are remembered. Mixed text is read by two voices: English words in English, Russian ones in Russian.
  • Long text is translated piece by piece and appended as each piece is ready, instead of appearing all at once ten seconds later.

Dictation

  • Three modes, each on its own key combination — ⌥Space, ⌥⌃Space and ⌥⌘Space by default, and every one of them can be rebound in Settings: click into the field and press the combination you want. Plain dictation types back exactly what you said, in Russian. Translated dictation turns Russian speech into natural English text as you speak. The secretary mode is for long sessions: it inserts text piece by piece, as each piece is recognised, instead of making you wait for the whole thing.
  • Dictation runs through a separate on-device speech model, chosen specifically because it keeps punctuation — Apple's own Russian dictation engine does not.
  • The secretary mode keeps writing into the field where the session started even if you switch to another app while still talking. It never translates: it always types back the words you actually said. Its voice commands ("point", "new paragraph" and the like) are an editable table in Settings — add your own, rename or remove the built-in ones.
  • A small panel shows the session as it happens. Drag it by its background and it stays where you left it, for both dictation modes.
  • A word-replacement dictionary fixes names, company names and terms that speech recognition keeps getting wrong, and one button (or its own hotkey) copies the last dictation to the clipboard.
  • The three dictation modes also show up as clickable items in the menu bar, next to the key combination that triggers them.

About privacy

Speech recognition and translation are done by the mechanisms of macOS itself, on the device. Transcriber opens no network connections of its own: no analytics, no update checks, nothing sent anywhere. The only thing that goes online is the system, once, when it downloads the speech recognition model.

The transcription history is kept as a plain file in your Application Support folder and is not encrypted. If you transcribe something that must not fall into the wrong hands, keep that file in mind.

HOW TO INSTALL

How to install

macOS will refuse to open it the first time

That is expected: the app is signed, but it has not been through Apple’s paid notarisation. Open it like this. First double-click the app in Applications — macOS will block it and show a window with two buttons: Done, and one that moves the app to the Trash. Press Done. That step is required: it is what makes the next one possible, and it stays possible for about an hour. Then go to System Settings → Privacy & Security, scroll down to Security, and press Open Anyway next to the app’s name. Enter your password when asked, then press Open. macOS remembers the choice — you do this once.

  1. Open the downloaded file and drag Transcriber into the Applications folder — it is shown right there in the same window.
  2. Launch Transcriber. No window will open: the app appears as a subtitle-bubble icon in the menu bar, at the top right. It will not be in the Dock.
  3. Click the icon and choose what to listen to. Under «Приложения» (Applications) are the apps that make sound — a browser, Zoom, a player; a note marks the one playing right now. Under «Микрофоны» (Microphones) are the microphones.
  4. Click «Начать» (Start). macOS will ask for permission to record the app's sound — or for the microphone, if that is what you chose — and for speech recognition. Allow both: without them there will be no subtitles.
  5. The first time, the system downloads the speech recognition model, so the subtitles appear a few seconds later. This does not happen again.
  6. The subtitle panel appears at the bottom of the screen. Drag it wherever suits you and stretch it by the bottom-right corner — the app remembers both.

If something seems off — no subtitles, dictation not typing anywhere — click the icon → «Разрешения…» (Permissions). It shows the live status of everything the app needs, with a button to fix whichever one is missing, instead of hunting through System Settings by hand.

To switch on translation under the cursor

This is a separate mode, and it does not switch itself on.

  1. Click the icon in the menu bar → «Настройки…» (Settings) → the section «Перевод под курсором» (Translation under the cursor) → switch on «Показывать перевод текста под курсором».
  2. Grant the Accessibility permission: System Settings (Системные настройки) → Privacy & Security (Конфиденциальность и безопасность) → Accessibility (Универсальный доступ) → switch on Transcriber. Without it the app cannot read text in another app's window.
  3. Go to any app, hold ⌥⌘ and point the mouse at English text. A quarter of a second later the translation appears beside it.
  4. If ⌥⌘ does not suit you, choose different keys and a different delay in the same settings.

If the translation still does not appear, click the icon in the menu bar → «Проверить текст под курсором…» (Check the text under the cursor). Move the mouse over text in different apps for twenty seconds — the app writes down what it sees under the cursor and then shows you where it put the recording.

To dictate text

Place the cursor where the text should go, press one of the three key combinations below, speak Russian, then press the same combination again to finish. All three (and every other hotkey in the app) can be changed in Settings by clicking into the field and pressing the combination you want — these are just the defaults:

  • ⌥Space — types back exactly what you said.
  • ⌥⌃Space — types the English translation instead.
  • ⌥⌘Space — a secretary mode for long dictation: it inserts text piece by piece as it is recognised, and needs the Accessibility permission from the previous section to keep writing into the right field if you switch apps mid-sentence.

A small panel shows what is happening; drag it by its background to move it out of the way. The same three modes also appear as clickable items in the menu bar.

WHAT’S NEW

What’s new

2.4 6 August 2026

Dictation and hover translation now work in Electron apps (Slack and similar) even when they don't.

  • Fixed dictation silently failing to insert text in some Electron apps (Slack confirmed) on some Macs: when accessibility can't identify exactly which app is focused, the app now sends the keystroke the same way a real keypress would travel — to whatever is actually focused — instead of giving up and leaving the text only in the clipboard.
  • Hover translation now wakes the target app's accessibility tree before reading the text at the cursor, the same way dictation already did. It didn't before, which is the likely reason it sometimes found a container with no readable text in Electron apps.
  • The three dictation modes and the secretary now lock each other out: starting one while another is recording does nothing, instead of running two sessions on the same microphone at once.
  • Added a "launch at login" switch (Settings → Subtitles & translation).
  • The dictation panel now says what it's doing — "Слушаю и перевожу…" for translated dictation, "Слушаю и записываю за вами…" for the secretary — and widens itself to fit instead of cutting the text off.
2.2 6 August 2026

A "Permissions" screen that tells you exactly what's granted, instead of guessing.

  • Added a Permissions window (menu bar icon → Permissions…) that shows the live status of everything the app needs — microphone, speech recognition, Accessibility, and app-audio capture — each with a one-click button to request it or jump straight to the right System Settings pane.
  • App-audio capture has no way to be checked in advance on macOS; the window now has a Test button that makes a real, harmless attempt right there, instead of requiring a full listening session just to find out whether it's allowed.
  • The menu bar now warns right at the top when something needed is missing, instead of failing silently later.
  • Fixed a misleading message during dictation: it used to guess "no Accessibility access?" even when Accessibility was already granted, sending people through pointless permission resets. It now checks the real state and logs the actual reason.
2.1 5 August 2026

Dictation now inserts reliably everywhere, and every hotkey in the app can be rebound.

  • Fixed dictation not inserting text in Electron apps (Slack, Discord, VS Code and others) while a non-Latin keyboard layout was active — the app now switches to a Latin layout for the instant it sends the paste keystroke, then switches back.
  • Every hotkey-bound action can now be rebound: click into the field in Settings and press the combination you want, instead of picking from a fixed list. Covers all three dictation modes, copying the last dictation, holding to see a hover translation, toggling hover translation on and off, and starting or stopping subtitles. A sequence already taken by another action is flagged right there.
  • The three dictation modes now show up as clickable items in the menu bar, next to the key combination that triggers them — click instead of remembering the shortcut.
  • Settings are split into three tabs — Dictation, Secretary, and Subtitles & translation — instead of one long scrolling form.
  • The secretary's voice-command list is now an editable table: add, remove or rename commands, instead of only switching the built-in ones on and off.
  • Faster startup and lower memory use: the speech-recognition and pause-detection models load once and are shared between dictation modes, instead of loading twice.
The whole journal →
Found a bug or something missing — tell us.

Something missing?

Tell us what broke or what you wish it did. We read all of it.

GET IN TOUCH

Have something worth building?

We take on a small number of partner projects each year. Tell us what you're making.

Or write directly: