Voice Editing

How do I edit text with my voice?

There are two ways to use Instructions Mode:

Tap-to-toggle (recommended for longer instructions):

Select text in any application
Tap the Instructions hotkey (default: Left Option + Shift) to start recording
Speak your command, then tap again to stop
An overlay appears streaming the AI response
Press Cmd+Shift+V to paste the result, or Cmd+C to copy

Hold (push-to-talk):

Select text in any application
Hold the Instructions hotkey and speak your command
Release — the AI response is pasted directly at the cursor (no overlay)

Hold mode is faster for quick edits. Web search is not available in hold mode.

What happens to the text I select?

The selected text is read via Accessibility APIs and sent along with your spoken instruction to Google Gemini. The raw audio is never sent — only the transcribed text.

How do I ask a follow-up question?

Press the Ask button in the overlay or hold the Instructions hotkey again. Previous Q&A pairs are shown collapsed in the overlay for reference.

Can I include clipboard contents in my instruction?

Yes. Say the word “clipboard” or “pasteboard” in your spoken instruction and MimicScribe automatically includes the current clipboard contents — including images — in the AI request. This works alongside selected text, so you can reference both.

Can Instructions Mode search the web?

Yes. Enable Web Search in Settings > Instructions. Responses will be grounded with live web results.

How do I change the AI model or prompt?

Settings > Instructions:

Provider — choose between Cloud or On-Device (if eligible). Cloud mode offers configurable model and thinking level options.
Instruction Prompt — editable system prompt with “Reset to Default”