Ubuntu’s Myna Speech-to-Text App Is Now Available as a Technology Preview

Today, Ubuntu maker Canonical announced the general availability of a technology preview of the Myna local speech-to-text dictation tool for Ubuntu users who find it easier to speak than to type. Myna brings AI-powered speech-to-text dictation to the Ubuntu Desktop using local models. While Myna prioritizes local speech recognition, feature releases will also support online models. However, the app doesn’t require an active Internet connection and processes audio in memory and discards it after use. For extra privacy, it only accesses the microphone when you explicitly activate dictation and doesn’t upload audio recordings to external services. Myna’s design is modular, which means that its features can evolve with future releases. Once local models are installed, everything is done on your machine. The initial release targets Ubuntu Desktop users on Wayland using the GNOME desktop environment. It supports several speech recognition engines, including Parakeet, Whisper, and FunASR, each with different trade-offs in language support, accuracy, and hardware requirements. The technology preview released today allows Ubuntu 24.04 LTS, Ubuntu 26.04 LTS, and Ubuntu 26.10 users on amd64 and amd64v3 systems to test-drive the speech-to-text app by running the commands below, one by one, in the Terminal app. Once installed, you can open Myna from Activities and follow the setup. sudo add-apt-repository ppa:canonical-desktop-team/myna sudo apt install myna-config For NVIDIA GPU users, Myna can also offer a GPU-accelerated engine. If you’re willing to try Myna, Canonical is interested in feedback on transcription quality, responsiveness, language and accent support, application compatibility, setup and onboarding, and anything else that feels awkward or unexpected. Image credits: Canonical