General Settings

General Settings

The General tab in the dIKta.me Control Panel settings houses the core application configurations, user interface language, and baseline integration behavior.

Behavior

  • Launch on Windows Startup: When checked (Autostart), dIKta.me will automatically launch and minimize to the System Tray every time you turn on your computer.
  • Play Feedback Sounds: Enables subtle audio cues (like soft clicks or chimes) when you start or stop a recording, letting you know the microphone is active without having to look at the Control Panel HUD.

Injection Settings

These options customize how the text lands in your target application.

  • Add space after injection: Automatically appends a single trailing whitespace after the dictated text is pasted. This allows for fluid, uninterrupted dictation across multiple sentences.
  • Append key after injection: Allows you to simulate a keyboard press immediately after the text lands.
    • None: Dictation stops precisely at the end of the text.
    • Enter: Automatically submits the text (perfect for chat applications like Slack, Teams, or Discord).
    • Tab: Automatically tabs to the next field (perfect for filling out spreadsheets or web forms).

Pipeline Overrides

  • Refine: Use Voice Instruction Mode: Changes how the Ctrl+Alt+R Refine hotkey operates.
    • Checked: You must press the hotkey, speak an instruction, and then release it (Voice Instruction mode).
    • Unchecked: Pressing the hotkey instantly processes the highlighted text using the system prompt without recording audio (Autopilot mode).
  • Ask Output Mode: Determines how the Ask pipeline delivers answers.
    • Clipboard and Toast: Answers are copied silently and shown in a Windows notification.
    • Clipboard Only: Answers are only copied to the clipboard.
    • Toast Only: Answers are only shown via notification.
    • Inject Only: Answers are pasted directly over your cursor.
  • Global Raw Mode: Forces all dictation pipelines to skip Artificial Intelligence processing and simply output whatever the Speech-to-Text provider recognized. Useful if you want absolute verbatim transcription with zero formatting.
  • Enable Streaming Dictation: Activates real-time dictation using a WebSocket if your STT provider (Deepgram) supports it. Replaces "Batch Mode" recording. Note: Streaming mode bypasses LLM system prompts and always outputs raw text.

Language Properties

  • UI Language: Controls the display language for the dIKta.me interface components like the settings menus and Control Panel buttons (e.g., English, Spanish). Changing this requires a restart.
  • Interaction Language: Controls the default expected spoken language sent to the AI processing pipelines (e.g., en, es, fr). It helps the STT engine prioritize your mother tongue.