SoundVibes (sv)

Open source voice-to-text for Linux

Overview

SoundVibes (sv) is an offline speech-to-text tool for Linux. It captures audio from your microphone using start/stop toggles and transcribes locally using whisper.cpp. No cloud, no latency, no subscriptions.

Quick Start

1. Install

curl -fsSL https://raw.githubusercontent.com/kejne/soundvibes/main/install.sh | sh

Or download manually from GitHub Releases.

2. Start the Daemon

sv daemon start

3. Toggle Recording

sv

Bind the toggle command to a hotkey in your desktop environment for hands-free use.

Documentation

• Website: https://soundvibes.teashaped.dev - Full documentation with installation guide, configuration reference, and troubleshooting
• Contributing: See CONTRIBUTING.md for development setup and build instructions

Quick Tips

Desktop Environment Setup:

• i3/Sway: bindsym $mod+Shift+v exec sv
• Hyprland: bind = SUPER, V, exec, sv
• GNOME/KDE: Add custom keyboard shortcut with command sv

Systemd Service:

systemctl --user enable --now sv.service

Requirements

• Linux x86_64
• Microphone input device
• Optional: Vulkan for GPU acceleration
• Optional: wtype (Wayland) or xdotool (X11) for text injection

See the website for detailed requirements and configuration options.

License

This project is licensed under the GNU General Public License v3.0 - see the LICENSE file for details.