Project Overview
VoxF9 is a Windows desktop application that enables real-time voice-to-text transcription with a single keystroke. Press F9 to start dictating, and your speech is instantly converted to text and inserted wherever your cursor is positioned. The entire process runs offline, ensuring complete privacy.
Key Features
-
Real-Time Transcription
- Live partial results while speaking
- Minimal latency between speech and text
- Automatic punctuation support
-
Complete Offline Operation
- No internet connection required
- Audio never leaves your computer
- No subscriptions or API costs
-
Simple Activation
- F9 hotkey to start/stop recording
- System tray integration
- Runs silently in background
-
Universal Compatibility
- Works in any application with text input
- Browsers, Office, IDEs, chat apps
- Seamless cursor position detection
-
Spanish Optimization
- Optimized for Spanish language
- Latin American accent support
- High accuracy for native speakers
Technical Stack
- Language: Python 3.12+
- Speech Engine: Vosk (offline recognition)
- Audio Capture: sounddevice
- Text Simulation: pyautogui
- System Integration: pystray, keyboard
- Platform: Windows 10/11
Installation
Two options available:
- Pre-compiled executable - Downloads ~1.4GB voice model on first run
- From source - Clone repository and install dependencies via pip
Use Cases
- Quick note-taking without typing
- Dictating emails and documents
- Accessibility for users with mobility limitations
- Hands-free coding comments
- Voice journaling and brainstorming
Open Source
VoxF9 is open source under MIT license. Contributions welcome!