VoiceType GUI

March 25, 2026 · View on GitHub

A graphical user interface for VoiceType that makes it easy to control voice dictation with Deepgram.

Features

  • API Key Management: Securely store your Deepgram API key in a persistent configuration file
  • Global Hotkey: Toggle dictation on/off using a customizable hotkey (defaults to F13)
  • Audio Feedback: Audible beeps when starting and stopping dictation
    • Start: Distinctive double beep (bright, high pitch)
    • Stop: Single lower pitch beep
  • Visual Feedback: Clean interface showing current status and recording state
  • One-Click Control: Start and stop dictation with a single button click

Installation

From Debian Package

sudo dpkg -i voice-keyboard_0.1.0-1_amd64.deb

From Source

cargo build --release
sudo cp target/release/voice-keyboard-gui /usr/bin/
sudo cp target/release/voice-keyboard /usr/bin/

Usage

First Time Setup

  1. Launch the GUI:

    voice-keyboard-gui

    Or find "Voice Keyboard GUI" in your application menu (Utility → Accessibility)

  2. Configure Your API Key:

  3. Set Your Hotkey (Optional):

    • The default hotkey is F13
    • You can change it to any key (e.g., F13, F14, etc.)

  4. Save Configuration:

    • Click "Save Configuration" to persist your settings
    • Settings are saved to: ~/.config/deepgram/voice-keyboard/config.json (Linux)

Starting Dictation

Method 1: Use the GUI Button

  • Click the green "Start Dictation" button
  • You'll hear a distinctive double beep (high-pitched) indicating dictation has started
  • The button will turn red and say "Stop Dictation"

Method 2: Use the Hotkey

  • Press your configured hotkey (default: F13)
  • You'll hear the same double beep
  • Works even when the GUI is minimized or in the background

Stopping Dictation

Method 1: Use the GUI Button

  • Click the red "Stop Dictation" button
  • You'll hear a single low-pitched beep indicating dictation has stopped

Method 2: Use the Hotkey

  • Press your configured hotkey again (default: F13)
  • You'll hear the same low beep

Audio Feedback

The GUI provides clear audio cues:

  • Start Recording: Double beep pattern

    • First beep: 1000 Hz (80ms)
    • Short pause: 50ms
    • Second beep: 1200 Hz (80ms) - brighter tone

  • Stop Recording: Single beep

    • 400 Hz (100ms) - lower, calmer tone

This makes it easy to know the current state without looking at the screen.

Configuration File

The configuration is stored in JSON format at:

  • Linux: ~/.config/deepgram/voice-keyboard/config.json

Example configuration:

{
  "api_key": "your-deepgram-api-key-here",
  "hotkey_code": "F13"
}

Security Note: The API key is stored in plain text. Protect this file:

chmod 600 ~/.config/deepgram/voice-keyboard/config.json

Permissions

The voice keyboard requires root privileges to create the virtual keyboard device. When you start dictation, you'll be prompted for your password via pkexec (PolicyKit).

This is required because:

  • The virtual keyboard needs to create a /dev/uinput device
  • This requires root access for security reasons
  • After creating the device, privileges are dropped to access your audio system

Troubleshooting

"Failed to start" Error

  1. Check your API key: Ensure your Deepgram API key is correct and saved
  2. Verify permissions: Make sure you have pkexec installed
  3. Test the core binary: Try running the voice-keyboard directly: 
    export DEEPGRAM_API_KEY="your-key-here"
sudo -E ./target/release/voice-keyboard --test-stt

No Sound / Beeps Not Working

  1. Check audio output: Ensure your system audio is working and not muted
  2. Volume levels: The beeps play at 30-35% volume - increase system volume if needed
  3. Audio system: Make sure PipeWire or PulseAudio is running

Hotkey Not Working

  1. Key availability: Ensure your keyboard has the configured key (e.g., F13)
  2. Key conflicts: Check if another application is using the same hotkey
  3. Try a different key: Some keys may not be available as global hotkeys
  4. Permissions: Some desktop environments require special permissions for global hotkeys

No Text Appearing

  1. Check microphone: Ensure your microphone is working and selected correctly
  2. Test audio input: Run voice-keyboard --test-audio to verify microphone levels
  3. Deepgram API: Verify your API key is valid and has credits
  4. Focus: Make sure the target application is focused and accepts text input

Build Errors

If you encounter build errors:

# Update Rust
rustup update

# Install required system dependencies (Ubuntu/Debian)
sudo apt install libasound2-dev

# Install required system dependencies (Fedora/RHEL)
sudo dnf install alsa-lib-devel

# Clean and rebuild
cargo clean
cargo build --release

Technical Details

  • GUI Framework: iced (native Rust GUI)
  • Global Hotkeys: global-hotkey crate
  • Audio Playback: rodio (for beep sounds)
  • Configuration Storage: serde + JSON
  • Platform: Linux (tested on Ubuntu with KDE Plasma/Wayland)

Command Line Interface

The core voice-keyboard binary also supports command-line options:

voice-keyboard [OPTIONS]

OPTIONS:
    --test-audio        Test audio input and show levels
    --test-stt          Test speech-to-text functionality (default)
    --debug-stt         Debug speech-to-text (print transcripts without typing)
    --stt-url <URL>     Custom STT service URL
    --voice-enter       Interpret "enter" at end-of-turn as Enter key
    --uppercase         Convert all typed text to uppercase

Contributing

This is a demonstration application for Deepgram's Flux API. For issues or improvements, please check the project repository: https://github.com/danielrosehill/deepgram-voice-keyboard

License

ISC License - See LICENSE.txt in the project root