SimplifyInstall apps

Speech to Text & Voice Dictation for Jira – User Guide

How to use Speech to Text for Jira: get started in 2 minutes, dictate in the issue panel, voice commands, the Library with templates, words and a signature, then post a comment or update the description.

Get started in 2 minutes

  1. Install Speech to Text from the Atlassian Marketplace (a Jira admin; free for up to 10 users).
  2. Open an issue in Chrome, Edge or Safari. Under its title, open the apps menu and choose Speech to Text — the panel appears above the activity.
  3. Click the round microphone and say a sentence; allow the microphone when the browser asks. Click it again to stop.
  4. Press Add a comment. Your words are on the issue, posted by you.

Overview

Speech to Text turns your voice into text in an editor on the Jira issue. Edit it if you like, then post it — as a comment, as the description (replace or append) or, in the Jira Service Management customer portal, as a comment on the request. Every change is made with your own Jira permissions.

Before you start

  • Browser: dictation uses the browser’s built-in speech recognition. It works in Google Chrome and Microsoft Edge on desktop and in recent Safari. Firefox has no speech recognition — there the microphone is disabled and you can still type.
  • Where your speech goes: your browser sends the audio to its own speech service — Chrome to Google, Edge to Microsoft; Safari may recognise on the device. The app itself never receives, sends or stores audio.
  • Microphone: the browser asks for access the first time you click the microphone.

Open the panel

  1. Install Speech to Text from the Atlassian Marketplace.
  2. Open an issue. Under its title, open the apps menu and choose Speech to Text. The panel appears above the activity.

The panel has its own editor and microphone: Atlassian does not let apps add a microphone to Jira’s own description or comment editor, or to the Create dialog.

Dictate, edit, post

  1. Choose the language: click the chip (“English (US)”) and search by language, country or code. Until you choose, the panel uses the language of your Atlassian profile; your choice is remembered in this browser.
  2. Optionally open ⚙ Options and turn on Stop after a pause — dictation then stops when you stop speaking. Otherwise it listens through pauses until you click the microphone again.
  3. Click the round microphone and speak. The line next to it says “Listening… click to stop”, with a small level meter. Each phrase becomes a paragraph, capitalised and ended with the language’s full stop; the words recognised so far show in grey under the text. Alt + Shift + R (macOS: Option + Shift + R) starts and stops dictation.
  4. Edit in the editor: headings, bold and italic, lists, links, code, tables, panels, status, date and colour. Type @ to mention someone, Link an issue to insert “KEY: summary” as a link.
  5. Post it with one button. Its main action is your choice — Add a comment (the default) or Append to description; ▾ next to it holds the other one and Overwrite description… (always asks first), and Main button sets which one the big button does. ⋯ has Clear the text… and the voice commands list.

When Jira confirms, the editor is cleared and the issue view refreshes to show the change (turn on Don’t refresh the page after adding in ⚙ Options to stay as you are and use Show it instead). When Jira refuses — no permission, the description is not on the edit screen, no connection — the message gives Jira’s reason and your text stays.

Unsent text is kept: leave the issue before posting and the text is back next time you open the panel on it in this browser (“Your unsent text was restored.”). Drafts are kept for two weeks.

Files

Attach files, paste a screenshot or drop files onto the editor. Each file appears as a link where the cursor is and waits under the editor. When you post, the files are uploaded as the issue’s attachments with your permissions; a comment with an image shows the image. If Jira refuses a file, nothing is written and your text and files stay.

Voice commands

Voice commands (on by default) work in English, Russian, Ukrainian, German, French, Spanish, Portuguese, Italian, Polish and Dutch — every regional variant. Commands lists what to say in the chosen language:

  • Punctuation — “comma”, “period”, “question mark”, “new line”, “new paragraph”…
  • Editing — “scratch that” takes back the last phrase, “undo” / “redo”, “clear all”, “bold that”, “bullet point …”, “insert template …”.
  • Actions — “add comment”, “append to description” and “overwrite description” press that button after a 3-second countdown; “cancel” stops it. Overwrite still asks before replacing.

Turn Voice commands off when you need these words as text (“trial period”).

The Library: templates, words and commands

Library next to the editor opens one panel (under the editor on narrow screens) with the tabs Templates, Words, Signature and Commands. Your last two templates sit as buttons next to Library — one click inserts. Type / in the editor and a few letters of a name to insert a template, your signature or a word where the cursor is.

  • Templates — texts you use often, with a search and level filters: Personal, Project and Site. Dictate or type, give it a name and save it; Insert puts it at the cursor; by voice say “insert template” and its name (“insert template bug report”). On a name clash the closest level wins: yours, then the project’s, then the site’s. Project administrators save project templates (Save for this project), Jira administrators site ones.
  • Words — words the speech service writes wrong: under Heard as what it writes (“j q l”), under Write as what you want (“JQL”). Dictated text gets the fix from then on; what you type is left alone.
  • Commands — what to say in the chosen language.

Templates and words are kept on Atlassian for your account, in every browser. Team words: Jira administrators also see Add for the team — team words fix everyone’s dictation (your own entry with the same name wins).

Signature

The Signature tab keeps your own signature — up to 500 characters, with Your name and Today’s date as fields filled in when it is used — with a preview.

  • Insert now, or say “insert signature” (in the ten command languages), puts it at the end of the text.
  • Add it automatically to comments and portal replies — off by default, never added twice, never to descriptions.
  • Administrators set a default for the project and for the site, used by people who have none of their own.

The customer portal

On a request in the Jira Service Management portal, the Reply by voice card offers the same dictation and editor — the language and ⚙ on its title line, the text first, then the microphone and Add a comment. The Library opens full screen, with Done:

  • Anyone who can see the request can use it — the reporter, participants, the organisation’s members and agents.
  • The comment is posted by you, with its files, as one comment. Bold, lists, headings, links, code and tables come through.
  • Where the service desk allows attachments, Attach files adds screenshots — also photos from a phone.
  • Mentions, issue links and team lists are not offered in the portal; the only voice action is “add comment”.

Languages

Over 60 languages and 101 regional variants — English in 12 regions, Spanish in 20, Chinese (Mandarin, Cantonese) and many more. Which ones work depends on the browser’s speech service. Open the language button next to the microphone and type a few letters to find one.

Permissions and data

  • Your permissions: if you cannot comment or edit the description, the app cannot either.
  • Permissions the app asks for: read and write Jira work (your comment or description, as you), read and write service desk requests (portal comments with files), read Jira users (the name in a signature) and app storage.
  • Stored on Atlassian: your templates, words, signature and the team’s entries, in the app’s Forge storage — not with Simplify. Usage counts per month (no text, no names) stay there too.
  • Stored in your browser: your language, the recording and voice-command settings and unsent drafts. On a shared computer, post or clear your text before you leave.
  • Outside Atlassian: only the audio, sent by your browser to its own speech service. The app’s code talks only to your Jira site.

See the privacy policy.

Troubleshooting

  • The microphone is disabled — the browser has no speech recognition (Firefox). Use Chrome, Edge or Safari.
  • “Microphone blocked” — press Allow microphone in the message, or allow it in the site settings next to the address, then reload.
  • Dictation stops by itself — Stop after a pause is on, or the speech service ended a long silence; click the microphone again.
  • The text is in the wrong language — choose the language before recording.
  • I don’t see a microphone in Jira’s editor — expected: use the Speech to Text panel, then Overwrite / Append to description or Add a comment.
  • “Speech to Text is not licensed for this site” — a Jira admin renews it in Manage apps.

Still stuck? Contact support.