Skip to content

Voice Input

Voice Input is one of Owlfy’s core features. Triggered by clicking the voice key, it provides an intelligent text input experience that goes far beyond simple transcription.

Click the voice key → Start speaking → Click again to end → Owlfy applies role polishing → Output appears at the cursor position.

Default Voice Keys:

  • Windows: Right Alt key or Mouse wheel
  • macOS: Fn key or Mouse wheel

You can customize the voice key in Settings.

Click voice key → Speak → Click again → Role polishing applied → Text output

This flow ensures your spoken words are refined before being entered, making voice input suitable for professional and formal contexts.

Choose from multiple roles to polish your voice input. This transforms casual speech into polished, context-appropriate text.

RoleEffectExample
Oral CorrectionConverts spoken language into formal written language”You know, I think we should probably…” → “I believe we should consider…”
Workplace ProfessionalRewrites casual speech into polished, high-EQ workplace expressions”Free for dinner tomorrow?” → “Would you be available for dinner tomorrow evening?”
Custom RoleCreate your own roles for personalized needsDefine any polishing rules you need
  1. Click “Role Management” in the voice input interface
  2. Click “Add Role”
  3. Set role name and polishing rules
  4. Save and use

Specify the output language to achieve real-time translation while recording.

Example:

  • Speak in Chinese: “明天一起开会讨论一下这个方案”
  • Set output language to English
  • Output: “Let’s meet tomorrow to discuss this proposal.”

This is perfect for:

  • Cross-language communication
  • Real-time interpretation during meetings
  • Quickly drafting messages in foreign languages

Map long text to short phrases for quick input.

How it works:

  1. Set a phrase mapping in Settings, e.g.:
    • Phrase: “Shipping address”
    • Full text: “123 Main Street, New York, NY 10001, (555) 123-4567”
  2. During voice input, say: “Shipping address”
  3. The full text is entered instantly

Common use cases:

  • Shipping/billing addresses
  • Company boilerplate text
  • Frequently used email signatures
  • Standard legal disclaimers
  • Complex code snippets

Privacy Guarantee: Snippets are stored locally with no network transmission.

  • Select your recording device
  • Test microphone levels
  • Ensure clear audio input

When enabled, system audio is automatically muted during voice input and restored afterward. This prevents background sounds from interfering with recognition.

Play subtle sounds when the voice bar appears and disappears, giving you clear auditory feedback on recording state.

You can speak punctuation marks directly:

"Hello comma are you there question mark" → "Hello, are you there?"
"One two three" → "123"
"Twelve thirty" → "12:30"

While recording, you can speak editing commands:

"Delete last sentence"
"Clear all"
"New line"

Voice input content is processed in real-time and not retained long-term. Your voice data is:

  • Used only for the current recognition session
  • Not stored on servers
  • Never used for model training