Whisperpad support readme

Whisperpad is a lightweight macOS app that provides fast, private, on‑device speech‑to‑text using local Whisper models. Everything is processed locally on your Mac—no cloud services, no accounts, and no data is ever sent anywhere.

How It Works

Whisperpad lets you dictate text into any app on macOS using a simple keyboard gesture:

1. Place your cursor inside any text field (Notes, Messages, Mail, Word, VS Code, etc.).
2. Double‑press Control to start recording.
3. Double‑press Control again to stop.
4. Whisperpad transcribes your audio and inserts the text directly into the active field.

That’s it—no switching apps, no separate windows to manage.

Settings

Open the Settings window to customize Whisperpad:

• Choose Whisper Model — Pick between a smaller, faster model or a larger, more accurate one. Press Reload Model after changing.
• Start/Stop Sounds — Enable or disable sound cues when recording begins or ends.
• Strip Notes — Remove text inside [brackets] or (parentheses) from transcripts.

Changes to filters apply instantly. Model changes apply after reload.

Features

• Fully offline transcription
• Native Apple‑silicon performance
• Insert text directly into any app
• Optional noise‑note cleanup
• Minimal, keyboard‑driven workflow
• No menu‑bar recording controls required

Requirements

• macOS with microphone access enabled
• Sufficient disk space for Whisper model files

Troubleshooting

• If no text appears, make sure the target app’s text field is active when you stop recording.
• If recording doesn’t start, check System Settings → Privacy & Security → Microphone and ensure Whisperpad has permission.
• If model changes don’t take effect, use Reload Model in Settings.

Privacy

Whisperpad processes all audio locally on your machine. No audio, text, or metadata is uploaded or transmitted.