HyperWhisper

A cross-platform desktop speech-to-text application with real-time transcription

---

About

HyperWhisper is a lightweight desktop application that provides real-time audio transcription using the Deepgram API. Record your voice, get instant transcriptions, and optionally auto-type the text directly into any application.

Features

• Real-time speech-to-text transcription
• Auto-type transcribed text directly into any application
• Audio recording with waveform visualization
• Recordings saved locally as WAV files
• Support for multiple audio input devices
• Dark theme UI
• Global keyboard shortcut support via D-Bus
• Works with HyperWhisper server or with Deepgram APIs

Installation

Download

Download the latest release for your platform from the Releases page.

Linux:

• .deb package for Debian/Ubuntu
• .rpm package for Fedora
• .AppImage for other distributions

nix build

Requirements

• Linux with PipeWire/PulseAudio for audio capture

• For auto-type feature: ydotool (Wayland) or xdotool (X11)

• Steps to enable auto-type on Linux distributions

  • make sure /dev/uinput is owned by root user and input group

    sudo tee /etc/udev/rules.d/99-uinput.rules << 'EOF'
    KERNEL=="uinput", MODE="0660", GROUP="input", OPTIONS+="static_node=uinput"
    EOF
    sudo udevadm trigger --name-match=uinput

  • create a ydotoold user service and enable it

    mkdir -p ~/.config/systemd/user/
    cat > ~/.config/systemd/user/ydotoold.service << 'EOF'
    [Unit]
    Description=ydotoold daemon
    
    [Service]
    ExecStart=/usr/bin/ydotoold
    Restart=always
    
    [Install]
    WantedBy=default.target
    EOF
    
    # Enable and start the service
    systemctl --user enable --now ydotoold.service

  • add your user to the input group

    sudo usermod -aG input $USER

• For Ubuntu/Debian:

  sudo apt install -y ydotool
  sudo dpkg -i hyperwhisper_0.1.0_amd64.deb

• For Fedora:

  sudo dnf install hyperwhisper-0.1.0-1.x86_64.rpm

• For NixOS:

  nix build

• Steps to enable auto-type on MacOS

  • Goto Settings -> Privacy & Security -> Accessibility
  • Add hyperwhisper here and enable it

• For MacOS:

  • you'll need rust and bun

    brew tap oven-sh/bun
    brew install bun
    
    brew install rust

  git clone https://github.com/hyperwhisper/app.git
  cd app
  bun tauri build

Usage

1. Launch HyperWhisper
2. Open Settings and configure your transcription service:
   • Hyperwhisper: Use the hosted service
   • Deepgram: Use your own Deepgram API key
3. Select your microphone
4. Click the record button or use the global shortcut
5. Speak and watch real-time transcription appear
6. Click stop to finish recording

Global Shortcut

You can trigger recording from anywhere using:

hyperwhisper transcribe toggle

or via D-Bus

dbus-send --session --type=method_call \
  --dest=dev.hyperwhisper \
  /dev/hyperwhisper \
  dev.hyperwhisper.toggle_recording

Bind this command to a keyboard shortcut in your desktop environment for hands-free operation.

Development

Prerequisites

• Rust (latest stable)
• Bun or Node.js
• Linux development libraries for Tauri

Setup

# Clone the repository
git clone https://github.com/hyperwhisper/app.git
cd app

# Install dependencies
bun install

# Run in development mode
bun tauri dev

Logo

bun tauri icon logo.png

Build

# Production build
bun tauri build

Build artifacts will be in src-tauri/target/release/bundle/.

Project Structure

app/
├── src/                    # React frontend
│   ├── components/         # UI components
│   ├── hooks/              # React hooks
│   └── App.tsx             # Main application
├── src-tauri/              # Rust backend
│   ├── src/lib.rs          # Core application logic
│   └── icons/              # App icons
└── package.json

Tech Stack

• Frontend: React 19, TypeScript, Tailwind CSS 4, shadcn/ui
• Backend: Rust, Tauri v2
• Audio: cpal (cross-platform audio)

License

GPLv3