Linux desktop
A hotkey, a tray icon, and your clipboard.
The desktop app does one thing well: you press Super+H, you talk, and the cleaned text is waiting on your clipboard. It is open source, so you can read exactly what it does with your microphone.
Clone and install
Installs the audio system packages, creates a virtual environment, and adds a desktop entry so Plainspoke shows up in your application menu.
git clone https://github.com/sebnado/plainspoke.git
cd plainspoke
./install.shSign in
Connects the desktop app to your account so transcripts land in the same history as the web app. Skip this if you only want local transcription.
./venv/bin/python main.py --loginHold the hotkey and talk
Press Super+H to start, press it again to stop. The cleaned text is on your clipboard by the time you switch windows.
What it needs
- Python
- 3.11 or newer
- Audio
- PipeWire, PulseAudio or ALSA — detected automatically
- Desktop
- GNOME, KDE, and wlroots compositors; X11 and Wayland
- Packages
- portaudio, libsndfile — installed by the script on apt and dnf
Local or cloud
The desktop app can transcribe entirely on your machine, with no account and no network. That mode is unlimited and always free — it costs us nothing, so it costs you nothing.
Signing in adds the cloud models, which are faster and more accurate, plus the shared history that shows up on your phone and in the browser. Cloud minutes are what the paid plans buy.
See what each plan includesmacOS and Windows
There are no native clients for macOS or Windows yet. The web app works on both, and on phones, with the same account and the same history — but it cannot bind a system-wide hotkey the way the Linux app can. We would rather tell you that than let you find out after paying.
Open the web app