Download and run
Open the installer and approve the Windows prompt. No administrator account is required.
Voice control for Windows
keyboard.wtf
Dictate anywhere, polish rough speech, and ask Jarvis to take useful action across your PC.
Download for Windows
Version 1.0 beta / Windows 10 and 11 x64 / 54.1 MB
Local speech models
Encrypted API key
Starts with Windows
Install once.
The installer sets up the app, registers startup support, and opens local settings while keyboard.wtf finishes downloading its speech and voice models.
Connect Gemini
Paste your Gemini API key in Set API Key, or use Get API Key to open Google AI Studio.
Start speaking
Use a global shortcut from any app. The setup page keeps model and startup status visible.
Three modes.
Literal when you need it, polished when you want it, and agentic when the computer should do the work.
Dictation
Fast, offline speech-to-text typed into the active app exactly as spoken.
Smart Writing
Removes fillers and false starts, fixes punctuation, and turns rough speech into clean writing.
Jarvis
Talk naturally, open apps and sites, prepare messages, manage windows, run workflows, and execute supported actions safely.
Speak from anywhere.
The same shortcuts work across Notepad, browsers, documents, email, chat, and your desktop.
Ctrl + Alt + DLiteral dictation
Ctrl + Alt + KAI-improved Smart Writing
Ctrl + Alt + QJarvis conversation and actions
Ctrl + Alt + XEmergency stop for any operation
Ctrl + Alt + ,Open local settings
Built to stay ready.
Settings, memory, speech models, and startup registration live on your PC. Sensitive actions still require confirmation.
First launch
Model downloads continue in the background and retry after restart. Gemini powers Smart Writing and Jarvis reasoning; offline transcription remains available locally.
Beta signing
This beta installer is not yet code-signed, so Windows may identify it as an unknown publisher. Download it only from this page.
SHA256 090A81F71788BBBFFCE1BD475EA34E7C1C071F1089CDEB27157F88A5D58FAF74