Tools2 min read

Myna for Ubuntu: local dictation is available to try

Canonical has opened the Myna preview for Ubuntu. What small businesses should check: compatibility, dictation quality and text handling in the destination app.

Prepared by the NADALI editorial team using AI

AI illustration: a worker dictates beside a laptop as sound waves become abstract text lines
AI illustration: Nadali / OpenAI · Original AI-generated output; use subject to applicable OpenAI service terms, not a stock/Creative Commons asset; no exclusive copyright guarantee. · 1440 × 960

Canonical opened the Myna dictation technology preview for Ubuntu on 12 October 2026. Ubuntu 26.10 followed on 15 October, with Myna available as an optional feature. This is an early version for testing, and it is not installed by default.

Bottom line. Teams already using compatible Ubuntu computers can test dictation for work drafts without sending audio to an external speech-recognition service. Do not assume time savings or accuracy in your language.

What is available

According to Canonical, once a model has been downloaded, Myna recognizes speech locally without an internet connection. The user activates a configured hotkey and speaks, and the text goes into the focused application. This is dictation, not voice control of the computer.

The preview is available on Ubuntu 24.04 LTS, 26.04 LTS and 26.10 for amd64 and amd64v3 with GNOME on Wayland. Trying it therefore does not necessarily require upgrading to the newly released operating system. Installation and initial setup instructions are in the official announcement.

Who may benefit and how to test it

The immediate audience is a small team that already uses Ubuntu and frequently types short texts. For example, someone could dictate a test draft of an internal note. This is an illustrative scenario, not a documented business case.

Our suggested approach:

  1. Ask the person responsible for IT to check compatibility and approve installing the preview on a test device.
  2. Select a few typical phrases without customer or confidential data. Test your language, accent, names and numbers.
  3. Compare dictation time, including corrections, with ordinary typing.
  4. Check where the text is inserted and read it before saving or sending it.

Where local processing ends

Canonical explains that audio is processed in memory and discarded after use. However, local recognition does not change the rules of the app receiving the text. Email or another work service may store or transmit that text according to its own settings.

Before using it for real work, define your data boundaries. Local transcription alone is not permission to dictate sensitive information.

Limitations

Myna remains under development. Language coverage, accuracy and hardware requirements depend on the engine; we have not tested Ukrainian dictation on your computer. We have not measured speed or time savings either. Wider adoption needs your own checks with work-related phrases and compatible applications.

Author: Nadali. Prepared with AI assistance. 

Sources and verification

  1. Canonical — Myna speech-to-text is now available as a technology preview (12 October 2026) ↗
    Source checked: 2026-10-19
  2. Canonical releases Ubuntu 26.10 Stonking Stingray (15 October 2026) ↗
    Source checked: 2026-10-19
  3. Ubuntu 26.10 release notes ↗
    Source checked: 2026-10-19
  4. Canonical — What’s new in security for Ubuntu 26.10? (6 October 2026) ↗
    Source checked: 2026-10-19