Skip to main content
Kalam Documentation

First-run setup

On first launch, Kalam opens a five-step setup wizard: Welcome → Access → Shortcuts → Engine → Try it. Every option can be changed later in Settings or Account.

Welcome step of the Kalam setup wizard showing account sign-in, Free option, and policy agreement. (Click to zoom)

The five setup steps​

  1. 1. Welcome — Policy and optional sign-in

    1. Read the Terms, Privacy Policy, and Beta Policy links, then check I agree to the Terms and Privacy Policy, and the Beta Policy.
    2. Choose Sign in or create an account (required only if you want a Pro trial, Pro license, or encrypted Sync), or skip sign-in and continue on Free.
    3. If your browser displays a verification code after sign-in, paste it into Paste the code from the website and select Paste code.
  2. 2. Access — Microphone and OS permissions

    1. Grant microphone access and any OS permissions listed (see Permissions by OS).
    2. Pick a microphone or keep the system default, then use Record → Stop → Play to verify you hear yourself cleanly.
    3. On Linux, check the Panel chip status row—if it reads Not added, follow Linux panel chip.
    4. On macOS, Microphone is required immediately; you can defer Accessibility, Input Monitoring, and Notifications with I'll do these later in System Settings, though global hotkeys require Input Monitoring before dictating outside Kalam.
    Access step on Linux showing microphone test controls and Panel chip status. (Click to zoom)
  3. 3. Shortcuts — Dictation keys and spoken languages

    1. Under Start dictating with, confirm or customize Hold to Dictate and Toggle Dictation.
    2. Under Languages you speak, keep your primary language as Default and use Add a language… if you dictate in multiple languages.
    OSDefault Hold to Dictate
    WindowsCtrl+Win
    macOSCtrl+Option
    LinuxCtrl+Alt

    Kalam waits 300 ms before recording starts so a quick tap does not start dictation by accident.

    Shortcuts & languages step showing Hold to Dictate, Toggle Dictation, and spoken languages. (Click to zoom)
  4. 4. Engine — On-device or Cloud

    On-device and Cloud speech engines are equal choices—Kalam sets no default for you.

    • On this device: Download a local model (SenseVoice Small, Whisper Base, Parakeet TDT 0.6B, or Moonshine Small). Audio stays on your computer.
    • Cloud: Select a provider tile, paste your API key, and verify it connects.

    Wait until the step reports your chosen engine is ready before advancing.

    Speech engine step offering equal Cloud and On this device tabs. (Click to zoom)
  5. 5. Try it — Test your first dictation

    1. Click inside the practice field on the Try dictation screen.
    2. Hold your Hold to Dictate shortcut (or press and hold the on-screen button).
    3. Speak a sentence and release. Your words appear directly in the field.
    Try dictation step before an engine is configured, prompting to choose a speech engine. (Click to zoom)

Change choices or re-run the wizard later​

StepWhere to change it later
Welcome / AccountSidebar → Account
Access / MicrophoneSettings → Voice Input
Shortcuts & languagesSettings → General / Dictation
Speech engineSettings → AI & Models & Dictation → Models

To walk through the five steps again without wiping your history, profiles, or keys, open Settings → General → Startup and select Re-run setup wizard.

Settings → General → Startup showing the Re-run setup wizard button. (Click to zoom)