This guide takes you from install to your first dictation — the same flow the app walks you through on first launch. Most people are talking to their apps within five minutes.
Install Mr.Speech
Download the app for your system and run the installer. On Windows, your browser may ask you to “Keep” the file twice — that’s normal for a brand-new app. On macOS, open the .dmg and drag Mr.Speech into Applications. On Linux, install the .deb (Debian/Ubuntu) or .rpm (Fedora/RHEL).
Sign in — or stay fully offline
Open the app and click Sign in. It opens your browser for a few seconds and then connects on its own — there’s no password to type. Your account is free during the beta and unlocks fast cloud transcription and AI cleanup.
Pick your dictation language
Choose the language you speak most often. Mr.Speech guesses it from your computer, so you usually just confirm. If you switch between languages a lot, pick Auto-detect — it keeps mixed-language dictation exactly as you said it.
Grant permissions
What’s needed depends on your operating system, and the app walks you through it.
Meet your two keys
Mr.Speech works with keys you hold. Hold the dictation key and talk — your words land wherever your cursor is. Hold the AI key (after selecting some text) and say how to change it. In setup you press each key once to see it light up; nothing records during that test.
Right Ctrl Dictation — hold and talk (Right Alt on macOS).
Alt + W AI edit — select text, then hold and say how to change it (Right ⌘ on macOS).
Download a local model (on-device mode only)
If you chose on-device mode, pick a local speech model to download: Tiny (77 MB, fast, runs on any PC), Base (148 MB, more accurate), or Small (488 MB, highest accuracy). Signed-in users can skip this — transcription runs in the cloud.
Your first dictation
Open any text box — an email, a chat, a code editor. Hold the dictation key, say a sentence, then release. Mr.Speech types it for you, cleaned up. That’s the whole loop, and it works in every app.
Try an AI edit
Select some text, hold the AI key, and say “make this more professional” or “shorten to three bullets.” Mr.Speech rewrites it in place. Red dictates, purple rewrites — the two keys you’ll use every day.
The five modes
Each mode has its own key and its own color on the floating pill, so you always know what’s about to happen.
Default shortcuts
Voice commands and the agent are opt-in — enable them during onboarding or in Settings.
Where to next
That’s the whole setup — now just talk. From here, explore how Mr.Speech turns what you say into a workspace: tune dictation in Make dictation yours, build a knowledge base in Notes and the graph, capture calls in Meetings, organize work in Plans, and go hands-free with the Agent & voice commands — all in the sidebar.
