Ubuntu users can install Myna’s documented configuration package from Canonical’s listed PPA and install the client through Snapcraft’s --edge channel. The client’s described dictation workflow requires GNOME on Wayland or Xfce on X11, plus a running IBus daemon. Canonical describes Myna as an experimental, offline speech-to-text project for Ubuntu Desktop.
Check the desktop requirements
Before trying dictation, check that your session is GNOME running on Wayland or Xfce running on X11, and that the IBus daemon is running. The Myna client README lists these requirements for its desktop app.
Install the documented packages
Canonical’s Myna project repository lists these commands for myna-config:
sudo add-apt-repository ppa:canonical-desktop-team/myna
sudo apt install myna-config
A SoftZone guide published October 12, 2026, names a different PPA, ppa:charles05/myna-config. Its installation guide and Canonical’s repository therefore give different PPA instructions; the repository lists ppa:canonical-desktop-team/myna.
Snapcraft separately lists this command for the Myna client:
sudo snap install myna --edge
The Snapcraft listing identifies --edge as the client’s installation channel.
Start dictation and insert text
The client README describes a custom keyboard shortcut for push-to-talk: press it to start dictation, then press it again to stop. Myna uses IBus to insert the transcript into the focused application. A text field must be focused for that insertion workflow.
How Myna handles audio
Snapcraft says the client captures audio and sends PCM to a local backend snap over a Unix socket. The inference snaps do not have microphone access; Snapcraft also says audio is not persisted and nothing leaves the machine. These details describe Myna’s local audio path in the Snapcraft listing.
Myna is experimental
Canonical warns that Myna and several of its features are in heavy development, and that major breaking changes may occur before a stable release. (Canonical’s project repository)
