Promote Sign in Sign up

Repository

SpeakAnywhere

Voice-powered productivity for Windows — dictate text into any app, and have any text read aloud with natural neural voices.

nicedreamzapp
Repository

README

Speak Anywhere

Voice-powered productivity for Windows - Dictate text anywhere and have any text read aloud with natural voices.

Features

Voice Dictation

Tap the microphone and speak - your words are typed automatically wherever your cursor is. Perfect for:

  • Writing emails and documents
  • Messaging and chat
  • Taking notes
  • Any text input field

Text-to-Speech

Copy any text to your clipboard, then tap "Speak Clipboard" to hear it read aloud using natural-sounding Microsoft neural voices. Great for:

  • Proofreading your writing
  • Accessibility
  • Multitasking while listening
  • Learning and comprehension

Additional Features

  • Speed Control - Adjust playback speed from 0.5x to 2.0x
  • Device Selection - Choose your preferred microphone and speaker
  • Offline Speech Recognition - Uses Vosk for privacy-focused, offline voice recognition
  • Desktop Shortcuts - Optional desktop and Start Menu shortcuts on first run
  • Portable - Run from USB drive without installation

Requirements

  • Windows 10 or Windows 11
  • Microphone (built-in or USB)
  • Speakers or headphones

Quick Start

  1. Download SpeakAnywhere.exe from Releases
  2. Download the Vosk model (see below)
  3. Run the executable

Option 2: Run from Source

# Clone the repository
git clone https://github.com/nicedreamzapp/SpeakAnywhere.git
cd SpeakAnywhere

# Install dependencies
pip install -r requirements.txt

# Download Vosk model (required)
# Place in _resources/vosk-model-small-en-us-0.15/

# Run
python speak_anywhere.py

Vosk Model Setup

Speak Anywhere uses Vosk for offline speech recognition. You need to download the model:

  1. Download from: https://alphacephei.com/vosk/models
  2. Get: vosk-model-small-en-us-0.15 (40 MB) or larger models for better accuracy
  3. Extract to: _resources/vosk-model-small-en-us-0.15/

How to Use

  1. Launch - Run SpeakAnywhere.exe or python speak_anywhere.py
  2. Wait for splash - The app loads the speech model (first time may take a moment)
  3. Setup options - Choose to create desktop shortcuts if desired
  4. Dictate - Tap the microphone icon, speak, and watch your words appear
  5. Listen - Copy text, tap "Speak Clipboard" to hear it read aloud
  6. Close - Click the X button to exit

Voice Commands

  • Say "new line" to press Enter
  • Say "period" or "comma" to insert punctuation

Building from Source

# Install PyInstaller
pip install pyinstaller

# Build executable
pyinstaller --onefile --noconsole --icon="_resources\SpeakAnywhere.ico" --name="SpeakAnywhere" --collect-all vosk --collect-all cv2 speak_anywhere.py

The executable will be created in the dist/ folder.

Project Structure

SpeakAnywhere/
├── speak_anywhere.py      # Main application
├── LICENSE.txt            # License agreement
├── README.md              # This file
├── requirements.txt       # Python dependencies
├── _resources/            # App resources (hidden folder)
│   ├── SpeakAnywhere.ico  # App icon
│   ├── splash_video.mp4   # Splash screen video
│   ├── splash_audio.mp3   # Splash screen audio
│   └── vosk-model-*/      # Speech recognition model
└── dist/                  # Built executable (after build)

Technologies Used

License

Copyright (c) 2024 Nice Dreamz LLC. All Rights Reserved.

This is proprietary software. See LICENSE.txt for details.

Author

Matt Macosko
Nice Dreamz LLC
[email protected]

Support

For issues and feature requests, please open an issue on GitHub.