Voice Commands (Supervoice)

Supervertaler includes Supervoice: voice dictation + Talon-style voice commands.

Where to find it

• Go to the Tools tab → 🎤 Supervoice.

Push-to-talk dictation (F9)

• Press F9 to start recording.
• Speak.
• Supervertaler transcribes what you said.

If voice commands are enabled, Supervertaler will first try to interpret your speech as a command. If nothing matches, the text is inserted as dictation.

Voice commands

Voice commands trigger actions like navigation and editing.

Examples of built-in phrases include:

• “next segment”
• “previous segment”
• “confirm”
• “copy source”
• “clear target”

You can add/edit/remove commands in 🎤 Supervoice.

Always-on listening mode

Always-on mode continuously listens for speech and automatically:

1. Detects when you start speaking
2. Records
3. Transcribes
4. Runs it as a command or dictation

This is configured and toggled inside 🎤 Supervoice.

Recognition engine

Supervoice supports two transcription modes:

• Local Whisper (offline): runs on your computer (slower, depends on your hardware)
• OpenAI Whisper API (online): faster and usually much more accurate for short commands (requires an OpenAI API key)

AutoHotkey (system commands)

Some voice commands can execute AutoHotkey scripts to control other applications (for example: memoQ or Trados).

INFO

If AutoHotkey is not installed, in-app commands still work, but system-level commands won’t.

Tips for accuracy

• Use a quiet environment and a consistent mic position.
• If recognition triggers too easily (or not at all), adjust Mic Sensitivity in 🎤 Supervoice.