Canonical is developing Myna, a local AI-powered speech-to-text tool for Ubuntu, set to debut in Ubuntu 26.10 (October). It works via a hotkey that activates a sandboxed inference process using open models like Whisper, Nvidia Nemotron, Parakeet, or Qwen3-ASR — all running locally with no cloud connectivity. Myna is a dictation tool only, not a voice assistant; wake words, voice commands, and continuous listening are explicitly out of scope. It will be distributed as a removable Snap package, and the project is currently in the architecture/planning stage on GitHub.
441 Impressions