Speech-to-Text Typer

A Python application that captures speech from your microphone and automatically types the transcribed text using Google Gemini API.

Features

• Record audio from microphone with manual control (Ctrl+C to stop)
• Transcribe recorded audio using Google Gemini API
• Automatic typing of transcribed text
• Multi-language support with automatic translation to English
• Toggle script for easy start/stop control
• Background operation with logging

System Requirements

Linux

Install these system packages before running the application:

Ubuntu/Debian:

sudo apt update
sudo apt install python3-dev portaudio19-dev xclip

Fedora/RHEL:

sudo dnf install python3-devel portaudio-devel xclip

Arch Linux:

sudo pacman -S python portaudio xclip

macOS

Most dependencies are built-in. You may need:

# Install Xcode command line tools if not already installed
xcode-select --install
# Install portaudio if needed
brew install portaudio

General Requirements

• Python: 3.13 or higher
• Microphone: Working microphone with proper system permissions
• Audio System: PortAudio for microphone input
• Display Server: GUI automation capabilities (built-in for most systems)
• Permissions: Microphone and accessibility permissions may be required

Setup

1. Install dependencies using uv:

   uv sync

2. Set up environment variables:

   cp .env.example .env

   Edit .env and add your Google API key:

   GOOGLE_API_KEY=your_actual_api_key_here
   

3. Get your API key from Google AI Studio

Usage

Direct execution:

uv run main.py

Recording will start immediately. Stop recording with Ctrl+C to transcribe and type the audio.

Toggle script (recommended):

./toggle_stt.sh

• First run: starts the application in background
• Second run: stops the application
• Logs are saved to /tmp/stt_typer.log

System-wide access:

sudo ln -s $(pwd)/toggle_stt.sh /usr/local/bin/toggle-stt
toggle-stt

Keyboard shortcut:

Add to Lubuntu keyboard shortcuts:

• Command: /full/path/to/your/project/toggle_stt.sh
• Key combination: your choice (e.g., Super+S)

How it works

The application records audio from your microphone until you stop it with Ctrl+C. It then uploads the audio file to Google Gemini API for transcription and automatically types the transcribed text. The system supports multiple languages and automatically translates non-English speech to English.