Linear

Linear Speech to Text: Dictate Issues, Specs and Updates

Most writing in Linear is explaining: what broke, why it matters, what done looks like. That's easier to say than to type. Here's how to dictate issues, comments and project updates in Linear, and how to keep them short enough that people actually read them.

  • Describe the bug out loud, then file it
  • Works in Linear's desktop app, browser and phone
  • Spoken steps become a numbered list

Also onMacWindowsiPhoneAndroid

Linear speech to text isn't built in: as of October 2026, Linear's docs don't describe a dictation button or any voice input. You can still talk your issues into Linear. LumeVoice types into the issue composer, comments and project updates in Linear's desktop app, in your browser and in Linear's phone apps, and cleans up your words on the way in.

Short answer

Does Linear have voice to text?

Not on its own. As of October 2026, Linear's help docs for issues, the editor, the mobile apps and Linear Agent don't describe any voice input. To dictate in Linear, use your device's dictation, such as the microphone on your phone's keyboard, Windows voice typing (Windows key + H) or Dictation on a Mac, or an app such as LumeVoice that types into any text field.

Dictating in Linear with the tools you already have#

Linear is made of ordinary text fields, so your device's own dictation works in it. As of October 2026:

  • iPhone. Tap the microphone on the keyboard. Apple says iPhone dictation is processed on the phone in many languages, with no internet connection required, which is a real privacy plus.
  • Android. In Gboard, tap the microphone at the top of the keyboard. You can say punctuation such as "period" and "new line."
  • Windows. Press Windows key + H in any text box. It needs an internet connection, and you can turn on auto punctuation in its settings.
  • Mac. Use the Dictation keyboard shortcut, or choose Edit → Start Dictation. In supported languages it adds commas, periods and question marks for you, and it stops after 30 seconds without speech.

These work well for a sentence or two. Longer explanations are where cleanup matters: dropping false starts, keeping only the corrected version of a sentence, and turning spoken steps into a list. That's what LumeVoice is built around.

How to dictate a Linear issue with LumeVoice#

  1. Download LumeVoice for Mac or Windows and sign in.
  2. In Linear, press C to open a new issue, or Option/Alt+C to start from a template.
  3. Type the title, or say it as one short sentence.
  4. Click into the description, hold fn (Mac) or Ctrl + Win (Windows), and explain: what happened, where, the steps, what you expected and what you saw. Let go, and the cleaned-up text lands in the description.
  5. Set the team, priority and labels as usual, and create the issue.

The same key works everywhere else in Linear:

  • Comments. Open the issue, click into the comment box and talk. Type @ when you want to mention someone, pick them from the list, then carry on.
  • Project updates. On the project overview, click the pencil icon on the latest update, choose On track, At risk or Off track, and say what moved, what's at risk and what's next. Updates can also be sent to a Slack channel, so they're worth getting right.
  • Docs and specs. Dictate a first draft of a project document, then edit it into shape.

Away from your desk, open Linear's app and use the Lume keyboard on iPhone or the Lume bubble on Android. Tap the center tab to create an issue, attach a screenshot from your camera roll, and describe the bug while it's fresh.

File the issue while it's fresh

LumeVoice turns a spoken bug report into a clean, punctuated Linear issue, with your steps as a numbered list.

  • One key in the desktop app or browser; a keyboard or bubble on your phone
  • Filler words out, self-corrections applied
  • 3,000 free words, no card required
Download for MacMac (macOS 11.5+), Windows 10 & 11, iPhone and Android

Tips for spoken issues people (and agents) can act on#

Keep the title short. Linear's own guide to writing issues favors short, plain titles that say what the task is. Say the title as one line, or type it, and save the detail for the description.

Talk in a shape. A bug report told in the same order every time is easy to read. An illustrative example:

You say: "Steps to reproduce: one, open a project with more than 500 issues. Two, filter by label. Three, scroll to the bottom. Expected: more issues load. Actual: the list stops and the spinner never goes away."

Lume types:

Steps to reproduce:

  1. Open a project with more than 500 issues.
  2. Filter by label.
  3. Scroll to the bottom.

Expected: more issues load. Actual: the list stops and the spinner never goes away.

Fill in templates by voice. If your team's issue templates have placeholder text, click each placeholder and dictate that part.

Say it, then trim it. Linear's guide also suggests writing only as much as the task needs. Talking makes it easy to say too much, so read it once and cut.

Write issues an agent can pick up. Linear lets you delegate issues to AI agents by assigning them, and Cursor's docs say you can work with its cloud agents from Linear. A spoken brief that covers the why, the constraints and what done looks like gives an agent the same context it gives a teammate. More on that for Cursor and Claude Code.

Let Linear Agent do the filing. Press Cmd/Ctrl+J to open a chat with Linear Agent, dictate your rough notes from a call, and ask it to turn them into issues. Linear's docs list creating and updating issues among the things it can do. Check what it files before you move on.

Keep names in the dictionary. Product names, codenames, customers and teammates belong in Lume's custom dictionary, so they come out the way your team spells them.

Your device's dictation vs LumeVoice in Linear#

Your device's dictationLumeVoice
PriceIncluded with your phone or computer3,000 free words, then $14.99/mo, $79.99/yr, or $249 lifetime
SetupNothing to installInstall the app and sign in
Where speech is processedVaries: iPhone handles it on the phone in many languages, a Mac can (check Keyboard settings), Windows voice typing needs internetIn the cloud; audio isn't stored after processing and is never used for training
CleanupPunctuation; on Copilot+ PCs, Windows fluid dictation also removes filler words in EnglishRemoves fillers, applies self-corrections, formats lists, fixes grammar
Across your devicesA different tool on each oneOne app on Mac, Windows, iPhone and Android

Swipe to see all columns

When your device's dictation is enough#

If you file a few short issues a week, your phone's keyboard microphone or your computer's dictation will do, and on an iPhone your speech can stay on the phone. That's a genuine advantage over LumeVoice, which processes speech in the cloud.

LumeVoice is worth it when Linear is where you do a lot of writing: long bug reports, specs, triage notes and weekly updates, plus the Slack threads and Notion docs around them.

Talk it through, then ship it#

Most of a good issue is context: what happened, why it matters, what done looks like. Say that part, keep the title short, and read it once before you create the issue. Your teammates, and any agent you hand it to, get what they need to start.

Frequently Asked Questions

Can LumeVoice create Linear issues for me?

Not by itself. LumeVoice writes down what you say; it doesn't click buttons, fill in properties or run commands. Press C to open a new issue, dictate the title and description, and create it as usual. If you'd like something to file issues from your notes, Linear Agent can create issues, and you can dictate those notes into its chat.

Do dictated lists keep their formatting in Linear?

They should. LumeVoice turns spoken lists into numbered or bulleted lists, and Linear's editor converts Markdown you paste into rich text, so a list of steps should arrive as a real list. Headings, code blocks and checklists are quicker to add with Linear's own shortcuts or its slash menu.

Can I turn a voice memo into a Linear issue?

Yes, on a computer. The LumeVoice desktop app transcribes audio and video files, such as m4a, mp3, wav, mp4 and mov. Open the file from the menu bar on a Mac or the system tray on Windows, or use Open With in Finder or the right-click menu in Explorer. Then put the transcript into the issue and trim it down.

Is LumeVoice private enough for internal bug reports?

LumeVoice sends your audio to the cloud for transcription and cleanup. The audio isn't stored after processing, nothing you say is used to train models, and your dictation history stays on your device. It needs an internet connection, and there's no team plan or admin console. If your company restricts third-party tools, check that policy before dictating about unreleased work.

Can I dictate in one language and file issues in another?

Yes. LumeVoice understands 100 languages and detects which one you're speaking. Set an output language and you can describe a bug in Portuguese, Hindi or German and have it typed into Linear in English for the rest of your team. Read it once before you create the issue, as you would with anything dictated.

Updated