Skip to main content
Kalam Documentation

What is Kalam

Kalam is voice dictation for your computer. Hold a shortcut, speak naturally, and release—your words are typed directly into whatever app you are using.

You choose where speech is recognized: on your computer or in the cloud, whichever you like. Kalam is free to use, while Pro adds AI polish, custom dictation profiles, Commands, and cross-device sync.

Placeholder until what-is-kalam-hold.webp is added under apps/docs/static/img/screenshots/.

Four things Kalam handles for you​

  1. Direct dictation in any app: Hold Hold to Dictate (or press Toggle Dictation) to speak into emails, editors, terminals, or chats without copy-pasting from a separate window.
  2. Equal choice of speech engine: Pick an on-device model (Sherpa-ONNX or Whisper.cpp) or connect your own cloud provider API key in Settings → AI & Models—with no forced default.
  3. Dictation profiles & Sense: Keep Plain dictation fast and verbatim, or create tailored profiles (Email, Notes, Coding) with AI polish, context rules, and automatic Sense routing.
  4. Commands & Ask Kalam: Speak prefixes like "New note…", rewrite highlighted text, run web searches, or ask Ask Kalam questions using your selection or clipboard.

On Windows, macOS, and Linux​

  • Windows & macOS: A compact floating pill tracks listening, transcribing, and error states near the edge of your screen. See Pill and overlay.
  • Linux: A lightweight panel chip lives directly inside your desktop panel (GNOME, KDE, Cinnamon, MATE, Xfce, or COSMIC). See Linux panel chip.