Basically superwhisper for Omarchy, Audetic is a voice to text application for Wayland/Hyprland. Press a keybind to toggle recording, get automatic transcription and inject text into the focused application/clipboard...

Quickstart Video

• View Documentation - Detailed guides and configuration

Quick Install

Linux ships a pre-built, signed binary:

curl -fsSL https://install.audetic.ai/cli/latest.sh | bash

The installer detects your platform and hands off to audetic install. Everything lives under $HOME — no sudo.

macOS builds from source — see the macOS Install Guide.

Linux

Copies the binary to ~/.local/share/audetic/bin/, installs a systemd user service at ~/.config/systemd/user/audetic.service, enable --nows it, waits for it to bind 127.0.0.1:3737, and opens the web UI in your browser. Pass --no-launch to skip opening the browser.

macOS

macOS builds from source — three commands once Xcode + brew install cmake ffmpeg are in place:

git pull
make macos-app
./target/release/Audetic.app/Contents/MacOS/audeticd install   # no sudo

Full walkthrough, permissions, local models, and troubleshooting: macOS Install Guide.

After installation:

1. Finish provider and ffmpeg setup in the web UI the installer opened (or visit http://127.0.0.1:3737/).
2. Add a keybind:
   • Hyprland: bindd = SUPER, R, Audetic, exec, curl -X POST http://127.0.0.1:3737/api/toggle
   • macOS: System Settings → Keyboard → Keyboard Shortcuts → Services / Shortcuts.app calling the same curl command.
3. Press the keybind to start/stop recording!

Web UI

The daemon serves a web UI at http://127.0.0.1:3737/ for onboarding, provider configuration, and browsing transcription history. The HTTP API lives under http://127.0.0.1:3737/api/* (e.g. POST /api/toggle, GET /api/status).

Configuration

Default config at ~/.config/audetic/config.toml. See Configuration Guide for details.

Provider CLI

Audetic ships an interactive helper so you can switch transcription providers without editing TOML by hand:

audetic provider show        # inspect current provider (secrets masked)
audetic provider configure   # interactive wizard (requires a TTY)
audetic provider test        # validate the stored provider

Transcribe Media Files

Transcribe audio or video files using the audetic cloud transcription service:

# Basic transcription (output to stdout)
audetic transcribe recording.mp4

# Specify language and output file
audetic transcribe meeting.mkv -l en -o meeting.txt

# JSON output with timestamps
audetic transcribe podcast.mp3 -f json --timestamps -o podcast.json

# SRT subtitle format
audetic transcribe video.mp4 -f srt -o subtitles.srt

# Copy result to clipboard
audetic transcribe voice-memo.m4a --copy

# Use custom API endpoint
audetic transcribe audio.wav --api-url http://localhost:3141/api/v1/jobs

Supported formats:

• Audio: wav, mp3, m4a, flac, ogg, opus
• Video: mp4, mkv, webm, avi, mov

Files are automatically compressed to MP3 before upload for efficient transfer. Files already in MP3 or Opus format are sent as-is. Use --no-compress to skip.

Options:

• -l, --language <LANG> - Language code (e.g., 'en', 'es', or 'auto' for detection)
• -o, --output <FILE> - Write transcription to file (default: stdout)
• -f, --format <FORMAT> - Output format: text (default), json, srt
• --timestamps - Include timestamps in text output
• --no-progress - Disable progress indicator
• -c, --copy - Copy result to clipboard
• --no-compress - Skip compression (send file in original format)
• --api-url <URL> - Override transcription API URL

Updates

Audetic includes an auto-updater plus manual controls:

audetic update

Uninstall — macOS

See the macOS Install Guide.

Uninstall — Linux

curl -fsSL https://install.audetic.ai/cli/uninstall.sh | bash

Use --dry-run to preview, or --keep-database to preserve transcription history. See Installation Guide for all options.

License

MIT