Dictation for Jira
Twenty tickets a week is not a heavy week for a lot of people, and almost none of the work in one is thinking. You already know what broke. What costs you the afternoon is typing it out: a title that reads like a title, a description somebody can act on, the steps in order, and then a screenshot dragged in afterwards to prove it. By the fourth one you are writing "see attached" and hoping.
Deixis is voice input for that. Hold Ctrl+Win, say the paragraph, let go, and clean punctuated
text is typed straight into the box you were looking at, with your clipboard handed back the way you
left it. Nothing is installed into Jira, because Deixis is not a Jira plugin and not an
extension for anything: it types wherever your cursor already is. The one key that dictates your
ticket also dictates your commit message, your email, your chat and your browser's address bar. One
voice input for the whole machine, not one per app — on Windows, Mac, iPhone and Android, with
speech recognition on the device every single time and audio that never leaves it.
The part worth reading twice is free mode. A bug report is mostly prose plus a picture, and free mode makes exactly that in one gesture: double-tap the key, keep talking while you drag a region of the screen, double-tap again, and you get one note with what you said and what you pointed at interleaved. Then a button of your own can file it and type the ticket key back at your caret.
Why Deixis
What it gives you in Jira, and what it will not do
Why Deixis in Jira
- Free mode is a bug-report machine. Double-tap
Ctrl+Winand Deixis keeps listening while you work. Tap the chord again for a small toolbar at the cursor: grab a region, annotate it, take a text selection, pick an element. Everything you said and everything you pointed at comes out as one note, pasted at the caret and saved as a file. That is the description and the evidence, produced while you were describing the problem rather than after it. Atlassian's own docs say a description can hold rich text, code snippets, images and tables — so the picture belongs in the body, and free mode is what puts it there. - Grabs you can mark up and redact.
Ctrl+Win+Sfreezes the screen so you can aim, then either quick-saves the region or opens the editor: arrows, boxes, a pen, a highlighter, text, numbered step badges, and a pixelate tool for the customer name that should not be in a ticket. Copy puts a real image on the clipboard, so it pastes straight into the description. Every save keeps a sidecar beside the PNG, so a grab stays re-editable rather than flattened for ever. - A recording with a link, for the bug you would rather show.
Ctrl+Win+Vrecords the screen you are on, with a small bar you can drag out of the way and pause. When a microphone was recorded, Deixis transcribes it on your own machine and writes it up with a title and a short summary, and Share a link hands you a URL anyone can watch with nothing to install. Sharing is the one managed feature, so that link needs a Deixis account of your own; it expires on its own and you can revoke it whenever you like. - Your own button, on the same palette. A custom button can run a program with your finished note and type back whatever it prints, straight at the caret — file the ticket, get the key back in the message you were writing. Or chain steps as plain settings: read the text off a screenshot, translate it, summarise it, post it, put the result where you want it. Nothing to script if you would rather not; the Buttons panel has ready-made ones a click adds.
- Signed webhooks to your own URL. Send every finished note, grab or translation to an endpoint you own, with an HMAC signature and a timestamp so your receiver can prove it came from you and refuse a replay. Delivery never blocks you: it retries on a background thread while your text is already pasted. Your URLs are yours — they do not pass through anybody's cloud.
- It punctuates for the field it is in. Jira's title box is one line and its description is prose, and the same sentence should not come out the same way in both. Deixis knows which it is typing into, because the name of the app and the name of the control ride along with your words. Never the window title, and never anything on your screen.
- Meetings, transcribed on this machine.
Ctrl+Win+Mrecords the microphone and, where the system allows it, the system output, transcribes both here, and writes a readable transcript into a folder you own. It tells the remote voices apart and remembers the ones you name, and the voiceprint that does it stays on the machine, is never uploaded, and can be deleted per person. Meeting notes — the summary and the action items — are the one optional step that leaves the machine, and only with an endpoint set. Half the tickets you write start in a call. - Everything is a file you own. Notes, transcripts, recordings, screenshots, your dictionary and every setting sit in one folder as ordinary files, with your dictation history behind its own tab and an export button. Back it up, grep it, uninstall tomorrow — it is all still there.
- Free, with no account, for ever. Dictation, free mode, grabs and annotation, screen recording, meetings, both speech models, the dictionary, all five vocabulary packs, your own buttons and your own webhooks. Deixis never asks you to sign in to dictate.
What it will not do
- It does not know your project's fields or your workflow. Deixis types words at a caret. It has no idea what your bug work item type requires, which component is which, or that this board wants a fix version set. A button you write does that — Deixis hands your script the note and pastes back what it prints. It is a door, not a Jira integration.
- It does not read your screen. Deixis sends the name of the app and the name of the field, never their contents, so it cannot see the ticket you have open and cannot fill anything in from it. What ends up in the description is what you said and what you deliberately pointed at.
- English only. The speech models are English models. If your team writes tickets in another language, this is not the tool for that yet.
- It will not transcribe a file you hand it. Deixis transcribes what it records — your dictation, your meetings, your screen recordings. There is no drop-a-file transcription in it.
- No Linux, and no PC older than 2013. The speech engine is compiled against an instruction set that arrived in 2013, so an older processor stops with an error rather than running slowly. Linux is not planned.
- It is not a keyboard replacement. Ticket keys, code snippets and paths are still faster typed. Deixis is for the prose half of a bug report, which happens to be most of it.
What you get
More than dictation. One key, and everything behind it.
Hold the key and press a letter: free mode, a grab, a meeting, a recording, a translation, your own buttons. All of it on your own device, all of it free, none of it behind an account.
Dictation, where your cursor is
Hold a key, speak, release. Clean, punctuated text lands in whatever has focus, on your PC, your Mac and your phone, with your clipboard handed back.
Free mode: talk and point
Narrate for as long as you like while you grab a region, pick a button on screen or take a selection. One note, what you said interleaved with what you pointed at, pasted where you finish and saved as Markdown.
Screen grabs, annotated
Freeze every screen at the keypress, drag a region, add arrows, boxes, text and numbered steps, and pixelate what is private. Saved as a PNG that stays editable.
Screen recording, shared as a link
Record the screen with your narration. The recording writes itself up with a title, chapters and a summary, and from your Deixis account it shares as a link anyone can open without one. The Loom you can stop paying for.
Meetings, no bot, names included
Your microphone and the other side, transcribed on your machine while the call runs. Deixis tells the voices apart and remembers the ones you name, and never uploads a voice.
Your own buttons and workflows
Add a button to the palette: it gets the note on stdin and types back whatever it prints. Send finished notes to any URL you own. Chain steps: read the text off a screenshot, translate it, summarise it, post it.
Field by field
The same voice, punctuated for wherever you are in Jira
Deixis knows which app and which field your words are going into — the name of the app and the name of the control, never the window title and never what is on your screen. A prompt, a message and a commit line therefore come out differently. Here is the browser, where the ticket is, kind by kind.
| Where you are | What you say | What Deixis types |
|---|---|---|
| A work item titlethe Title box on the create dialog — one line, and always required | “um the release poller reads the feed twice on a cold start” | Release poller reads the feed twice on a cold startA one-line box gets a line: the filler goes, it takes a capital, and it is not handed a full stop it does not need. |
| A descriptionthe long body under the title, where the detail goes | “the release poller reads the feed twice on a cold start it only happens when the cache is empty so we see it after every deploy” | The release poller reads the feed twice on a cold start. It only happens when the cache is empty, so we see it after every deploy.The same sentence in a long body comes out as sentences, because Deixis knows it is typing into a description and not into a one-line box. |
| Steps to reproducethe numbered part of the description, dictated in one breath | “deploy the worker clear the cache then watch the feed log basically you get two requests instead of one” | Deploy the worker, clear the cache, then watch the feed log: you get two requests instead of one.Steps said carefully come back punctuated as steps. The filler goes and not one of your words is rewritten. |
| A commentthe comment box at the bottom of a work item | “erm i can't reproduce this on staging can you say which shad kin version you're on” | I can't reproduce this on staging. Can you say which shadcn version you're on?Your dictionary fixes shadcn on this machine before anything is sent anywhere, and a question gets a question mark. |
| A status updatethe comment you leave when you move a work item to the next column | “moving this over the fix is on the branch and the test covers the cold start case it needs a review from someone on the platform side” | Moving this over. The fix is on the branch and the test covers the cold start case. It needs a review from someone on the platform side.Three thoughts said in one breath come out as three sentences, which is the difference between a note somebody reads and one they skip. |
Jira facts from create, describe, attach, mobile, jiraai, rovodesktop, rovomobile and rovochat, read on 7 September 2026. Everything about Deixis on this page is from its README for the current release.
Worth knowing about Jira
- Atlassian's own create dialog puts the two fields side by side: “Type a title for the work item. This field was previously labelled Summary,” and a description that “expands vertically to accommodate long descriptions, code blocks, table edits, and pasted images.” One is a line, the other is prose — which is exactly the distinction Deixis punctuates for.
- The order on the form is the order of the work. Atlassian's create dialog asks for the title first and puts the description beneath it, so every ticket you file starts with a line you have to compose before you reach the detail.
- Atlassian says “A description can include rich text, code snippets, images, tables, synced blocks, and more,” and that you can drop or attach files in comments and other long text fields — so a picture belongs in the body of the ticket, not only bolted on underneath.
- There is plenty of room for evidence: Atlassian's docs say you can add up to 2000 attachments to a work item.
- Jira's mobile apps for Apple and Android let you “create work items, assign work, check on tasks, make comments, and much more,” and Atlassian pitches them for when you need to record tasks from a meeting or respond on the fly. Nothing on that page mentions voice, dictation or a microphone.
- Atlassian's page for AI in Jira lists work create, work breakdown, instant context, agents and a browser extension. Voice, dictation and speech input appear nowhere on it: the AI writes for you, it does not listen to you.
- Atlassian does have dictation, but it belongs to Rovo rather than to a Jira field. Rovo Desktop — Windows and macOS — lists “Use dictation to create and edit prompts” among the things the microphone permission is for, and the words go into Rovo's prompt box, not into the title, description or comment you are looking at.
- The Rovo mobile app on iOS and Android says you can “capture context with photos or voice,” including uploading a photo of a whiteboard and asking Rovo to create Jira work items from it. Again it is voice at the assistant, not voice at the caret.
- Rovo Chat itself reaches into Jira, Jira Service Management, Jira Product Discovery, Confluence and Bitbucket, and is also available through Rovo Desktop, the Rovo mobile app, a browser extension, Slack and Microsoft Teams.
Five minutes
Setting it up
About five minutes, and nothing goes into Jira at all.
Install and fetch the model
One installer, nothing else to install alongside it. The first launch shows a single Download button for the speech model — about the size of a short video, fetched once, and checked against its published size and hash before it is kept. After that you are offline unless you choose otherwise.
The gesture
Hold Ctrl+Win, speak, release: the text lands where your caret is. Double-tap it instead for
free mode, which keeps recording until you double-tap again — that is the one for a bug report,
because it lets you talk while you grab. Hold the chord a moment longer and a small palette appears
with a letter per mode: S grabs and annotates, V records the screen, M records a meeting, T
translates a selection, D puts the word under your cursor into your dictionary. If the chord
fights something in your setup, Settings takes a new one as you press it.
Teach it your words
The single biggest accuracy win, and it costs nothing in speed. Your project keys, your component
names, the service that keeps coming out as two words: select one anywhere — in a ticket, in a
README — press Ctrl+Win+D, and it is in your vocabulary. Corrections for names it keeps mishearing
are unlimited and applied on your own machine after the transcription, before anything else sees the
words. Five packs ship in the box and the software one is on by default.
Decide what may leave, once
Settings → Cleanup is the only privacy decision here, and its default is the quiet one: rules tidy your words on this machine unless you asked for a transform or dictated a long stretch with no punctuation. Set it to never, or leave the endpoint empty, and dictation never touches the network. Point it at a model running on your own machine and the better prose stays local too. Whatever you pick, your audio is not part of it — that never leaves, and there is no setting that makes it.
File the ticket without leaving the box
This is the ten minutes worth spending once. Add a button to the palette that runs a small script with your finished note; your script gets the note as JSON, arguments go straight to the process as a list so nothing needs escaping, and whatever it prints is typed back at your caret. So "the release poller reads the feed twice on a cold start", narrated over a grab of the log, becomes a work item — and its key appears in the message you were already writing. The first run shows you the exact command line and does nothing until you approve it; editing the command asks again.
If you would rather not run anything locally, point a signed webhook at a URL you own instead and have your own service create the work item. Same note, same signature, no script on your machine.
Before you ask
Questions people search for
Does Jira have voice input of its own?
Not into a ticket. Atlassian's pages for creating work items, for the mobile apps and for AI in Jira describe titles, descriptions, slash commands, agents and summaries, and mention no voice, dictation or microphone anywhere. Atlassian does have dictation, but it belongs to Rovo: Rovo Desktop on Windows and macOS asks for microphone permission so you can use dictation to create and edit prompts, and the Rovo mobile app says you can capture context with photos or voice. Those put your words in the assistant's prompt box, not in the description you are looking at.
Can I dictate straight into the title and the description?
Yes, and they come out differently on purpose. The title box is one line, and Atlassian's own description field expands vertically for long text, code blocks and pasted images. Say the same sentence into each and Deixis gives the first a line and the second sentences, because the name of the field travels with your words.
How do I get a screenshot into the ticket at the same time?
Double-tap Ctrl+Win, describe the bug while you drag a region of the screen, then double-tap to
finish. You get one note with the picture in it, pasted where your caret is and saved as a file.
Annotate or redact the grab first if it needs it, and Copy in the editor puts a real image on
the clipboard for the description body.
Do I need an account or a subscription?
No. Dictation, free mode, grabs, screen recording, meetings, both speech models, the dictionary, your own buttons and your own webhooks are free with no sign-in at all. The optional cloud cleanup is free too if you bring your own provider key or point it at a model on your own machine. Paying only buys not having to set that up. Sharing a recording as a link is the one thing that needs a Deixis account.
Does it work outside Jira too?
Everywhere. Deixis is not an extension for a particular app: it types at the caret in whatever has focus, so your ticket, your pull request, your terminal, your email, your chat and the address bar all take the same key. What changes between them is the punctuation.
Will it work offline?
Completely. With no cleanup endpoint set, dictation, free mode, screen grabs, screen recording and meeting recording and transcription all run with no network at all. Two calls a fresh install makes on its own: the update check, and — when you open the Settings window — a check for an announcement from us. Both send nothing but the request.
Keep reading
Related pages
Voice input for your product, without the audio leaving your user
Put voice input in your web or desktop app in about ten lines. Recognition runs on your user's machine, so no audio reaches you, us, or any cloud.
Offline dictation on Windows
Dictation that works with no internet: speech recognition on your own device, free, no account. Which Windows speech-to-text tools genuinely run offline.
Dictation for Slack
Deixis dictates into Slack on one key — channel replies, threads, DMs and canvases. Speech recognition on your own device, no account, free.
Dictation for Microsoft Teams
Deixis dictates into Microsoft Teams on one key, and records and transcribes your calls on your own machine. No bot in the call, no account, free.
Dictation for Outlook
Deixis dictates into Outlook on one key — subject lines punctuated as subject lines, bodies as prose. Speech recognition on your device, no account.
Dictation for Claude Code
Deixis dictates into Claude Code on one key — terminal, editor, desktop app or browser. Speech recognition on your own device, no account, free.
Hold a key. Speak. Keep working.
Deixis is free, needs no account, and runs on your own device. Everything on this page is in the current release.
v1.2.2 · 64‑bit Windows 10/11 · macOS 13 on Apple silicon · 51 MB · nothing else to install · also for Windows
Facts about Jira checked on 2026-09-07 against atlassian.com. Deixis facts are from its README for the current release. Something out of date? Tell us.