Canonical has launched Project Myna (named after the myna bird, known for mimicking human speech), an Ubuntu Desktop initiative that integrates speech-to-text dictation into the operating system.
The first release is scheduled for Ubuntu 26.10 (Stonking Stingray), and plans are to make it a core part of the Ubuntu Desktop experience
This feature is being developed as a native desktop dictation tool. Users can press a keyboard shortcut, speak, and have the transcribed text inserted into their current application. The initial version will also offer visual feedback during dictation.
For the initial release, Canonical is keeping the scope narrow. Project Myna will not serve as a voice assistant, voice-command system, translation tool, or desktop-control feature. Automatic language detection is also excluded. The focus is only on reliable desktop dictation.
Myna uses AI speech recognition models that run locally on the user’s machine and require no internet connection after model installation. The initial release targets Ubuntu Desktop on Wayland, with GNOME as the primary supported environment, with plans to add support for additional desktop environments in the future.
Access to the microphone will be granted only when dictation is activated. Audio is processed in memory and discarded after use, with no recordings uploaded to external services. The architecture is modular, with speech recognition managed separately from user interaction, dictation management, and text injection.
After Ubuntu 26.10, Canonical will continue to improve desktop integration and seek ways to make dictation more natural and accurate. Priorities will be guided by feedback from early users and the wider community.
For additional details, see the announcement.
