Canonical has released a preview version of the Myna speech recognition system and the text input tools based on it, which can be integrated into Ubuntu. Ready builds are available for Ubuntu 24.04 LTS, Ubuntu 26.04 LTS, and Ubuntu 26.10 in packages for amd64 and amd64v3 architectures. It supports operation in GNOME environments that use the Wayland protocol. For speech recognition, users can choose from AI models Whisper, Parakeet, or Funasr, which run locally and support various languages.
An extension for GNOME Shell, called Myna Shell, has been prepared for desktop integration. Voice input is activated by the Super+J key combination, after which text can be dictated aloud, and pressing Super+J again ends the dictation session. The dictated text is inserted into the current application by simulating keyboard input.
The configurator has added a section for setting up the activation key combination, indicating the start/stop of recognition, auto-disconnecting after a certain period of silence, and a text substitution mode (either as spoken or all at once after finishing dictation), as well as choosing the AI model. An engine for hardware acceleration of speech recognition models on systems with NVIDIA GPUs is provided (other acceleration options will appear later).
Installation is done via the PPA repository:
sudo add-apt-repository ppa:canonical-desktop-team/myna
sudo apt install myna-config
Source: opennet.ru
