sttapp

System-wide Speech-to-Text Desktop Application

A fast, lightweight desktop app that transcribes speech to text using OpenAI-compatible APIs. Features global hotkey support, automatic clipboard copying, and seamless paste-to-active-window functionality.

Installation

Pre-built Binaries

Download the latest release for your platform from GitHub Releases:

• Windows: .msi or .exe installer
• macOS: .dmg or .app
• Linux: .AppImage, .deb, or .rpm

Build from Source

Prerequisites

• Rust (latest stable)
• Deno runtime

Build Steps

git clone https://github.com/lmtr0/sttapp.git
cd sttapp
deno task build
deno task tauri build

Binaries will be available in src-tauri/target/release/bundle/.

Configuration

First-Time Setup

On first launch, the settings window opens automatically. Configure your API credentials to get started.

API Configuration

The app requires an OpenAI-compatible API:

• API Key: Your API key (e.g., sk-... for OpenAI)
• Base URL: API endpoint (default: https://api.openai.com/v1)
• Model: Speech recognition model

Preset Models

• whisper-1 - OpenAI Whisper
• whisper-large-v3 - Groq Whisper v3
• whisper-large-v3-turbo - Groq Whisper v3 Turbo

Custom models can be configured for other OpenAI-compatible APIs.

Environment Variables

Alternatively, configure via environment variables:

export OPENAI_API_KEY="sk-your-key-here"
export OPENAI_BASE_URL="https://api.openai.com/v1"  # optional
export OPENAI_MODEL="whisper-1"  # optional

Supported Providers

• OpenAI - Default, uses whisper-1
• Groq - Fast inference, uses whisper-large-v3 variants
• Any OpenAI-compatible API (custom base URL)

Test Connection

Use the "Test Connection" button in settings to verify your API credentials before saving.

Usage

Quick Start

1. Launch sttapp
2. Configure API settings
3. Press F8 to start recording
4. Press F8 to stop and transcribe
5. Transcript auto-copies to clipboard and pastes into active window

Keyboard Shortcuts

Shortcut	Action
F8	Start/stop recording, normal paste
Shift+F8	Start/stop recording, paste as plain text

System Tray

Right-click the tray icon to access:

• Start Recording - Begin audio capture
• Stop Recording - End capture and transcribe
• Settings - Open configuration window
• Quit - Exit application

The tray icon changes state during recording.

Workflow

1. Press F8 (or use tray menu) to start recording
2. Speak into your microphone
3. Press F8 again to stop
4. Audio is encoded to FLAC and sent to the API
5. Transcript is copied to clipboard
6. Transcript is automatically pasted into the active window
7. Main window auto-hides (if not focused)

Window Behavior

• Positioned at bottom-center of screen
• Auto-hides after successful transcription
• Can be closed at any time without affecting recording

Development

Prerequisites

• Rust (latest stable)
• Deno runtime
• Platform-specific build tools:
  • Windows: Microsoft Visual Studio C++ Build Tools
  • macOS: Xcode Command Line Tools (xcode-select --install)
  • Linux: build-essential, libgtk-3-dev, libwebkit2gtk-4.0-dev

Development Commands

# Start frontend dev server
deno task dev

# Run Tauri in dev mode with hot reload
deno task tauri dev

# Type check
deno task check

# Build frontend
deno task build

# Build production binaries
deno task tauri build

Project Structure

sttapp/
├── src/                    # SvelteKit frontend
│   ├── routes/            # SvelteKit routes
│   │   ├── +page.svelte   # Main recording window
│   │   └── settings/      # Settings page
│   └── lib/               # Shared utilities
│       └── config.ts      # Configuration management
├── src-tauri/             # Rust backend
│   ├── src/
│   │   ├── main.rs        # Entry point
│   │   └── lib.rs         # Core app logic
│   └── tauri.conf.json    # Tauri configuration
└── static/                # Static assets
    └── audio-processor.worklet.js  # Audio capture worklet

Key Components

• Audio Capture: Uses AudioWorklet for efficient 16kHz mono capture
• FLAC Encoding: Client-side encoding via libflac.js
• API Integration: OpenAI-compatible transcription endpoint
• Global Shortcuts: F8 and Shift+F8 via Tauri global-shortcut plugin
• Auto-paste: Uses enigo for cross-platform keyboard simulation

Platform Notes

• Linux/Wayland: Microphone permission handled programmatically for WebKitGTK
• Linux/X11: Requires DISPLAY environment variable
• Windows: No console window in release builds

Contributing

Contributions are welcome! Here's how to get started:

Development Workflow

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Make your changes
4. Test thoroughly using deno task tauri dev
5. Commit your changes (git commit -m 'Add amazing feature')
6. Push to the branch (git push origin feature/amazing-feature)
7. Open a Pull Request

Code Style

• Frontend: TypeScript with Svelte 5 runes ($state, $derived)
• Backend: Rust with standard Rust formatting
• Commits: Clear, descriptive commit messages

Reporting Issues

Report bugs or request features via GitHub Issues.

Please include:

• Operating system and version
• Steps to reproduce
• Expected vs actual behavior
• Screenshots (if applicable)

License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.