Voice to text in Confluence

Does Confluence have voice-to-text?

No. There's no dictation button in the Confluence editor, no microphone in the toolbar that types what you say, no setting buried in a menu that turns speech into a page. I went looking, the way you probably did. It isn't there. Atlassian has shipped a lot into that editor over the years, live docs, inline comments, whiteboards, databases, but voice input isn't one of them.

So the dictation has to come from outside Confluence. Two places it can come from: your operating system's built-in voice typing, or a dedicated desktop app that types into any text field. Both work in the Confluence page editor, the comment box, and the live-doc surface, because as far as your computer is concerned, you're just typing. Both are below.

One thing to set aside first. "Transcribing a recorded meeting into a Confluence page" is a different job from "dictating a page as I write it." This guide is about the second one, talking your draft into the editor. If you want a recording turned into notes, that's a transcription tool, not dictation, and it's a separate category.

Writing a long doc by talking

The case for dictating Confluence pages is mostly about the long ones. Speaking runs about 145 words per minute; typing for most people is closer to 40. That's roughly three and a half times faster. On a one-line comment it makes no difference. On a runbook, a postmortem, a project brief, an onboarding page, the kind of doc you keep putting off because it's an hour of typing, the gap is the difference between writing it today and writing it never.

And talking changes how the draft comes out. When you type, you edit as you go, which is slow and tends to strangle the first version. When you talk, you get the whole messy thought down, then clean it up. For documentation specifically, where the hard part is usually getting the explanation out of your head and onto the page, that ordering helps. The boring truth is most docs don't get written because the typing is a wall. Talking lowers the wall.

The fastest way: a system-wide hotkey

Here's where a dedicated app changes the math. Whisper by Remskill is not a Confluence app, a Marketplace add-on, or a browser extension. It's a desktop app that works like a keyboard: press a hotkey, speak, and the transcript is typed at the cursor, in any app, the Confluence editor included. It works the same in Confluence open in a browser and in the Confluence desktop app, because your computer just sees text being entered.

Setup is short:

That's the whole loop. The text appears in the editor, you read it, you keep going. For a long page you talk in chunks, a paragraph at a time, the same way you'd think it.

Speak, and the page fills in

Once it's running, the experience is unremarkable in the best way. You put your cursor in the Confluence body, hold the key, talk, let go. A second or so later the text is sitting in the editor as if you'd typed it. No copy-paste from a separate window, no recording to transcribe later, no add-on to authorize against your Atlassian account.

Because the local transcription runs on your machine (pure-Rust, no Python sidecar, no server in the loop), it works offline, and the audio for whatever you're dictating never leaves your laptop. That matters for work docs in particular. A page might describe an internal system, a customer, an incident; with local mode, none of that audio goes to anyone's server to become text. Windows' own built-in voice typing needs an internet connection; Whisper's local mode does not.

Turning a spoken draft into clean prose

Spoken language is messy. You say "um," you restart sentences, you trail off into a noise that means "you know what I mean." For a Confluence page that other people read, you don't want that on the page. Whisper has an optional AI enhancement step that trims filler and tidies the phrasing before it types. So "uh, so the, the deploy script, it kind of, it needs the env var set first, otherwise it, yeah it just fails" becomes "The deploy script needs the env var set first, otherwise it fails." For documentation, that cleanup is the difference between a transcript and a draft.

That cleanup runs locally through Ollama, free, on your own machine. Pro users can route it through the cloud instead, but the filler-cleanup benefit doesn't require Pro; it's there in the free local pipeline. You can also turn it off and type the raw transcript, then edit in Confluence the normal way, which for a quick comment is often the right call.

What the full Whisper app looks like

The hotkey is the part you'll use most, but there's a settings surface behind it. You pick your transcription engine: Whisper models, whose multilingual variants cover 99 languages and the English-only .en builds cover exactly one, or NVIDIA's Parakeet, about 600 MB, 5 to 10 times faster than Whisper on CPU, covering English plus 24 European languages. You set your hotkey, manage history, and save presets, handy if your docs lean on the same product names or jargon a lot. None of that is required to dictate one page. It's there when you want to tune.

Works in the browser editor and the desktop app

People reach Confluence two ways: in a browser tab and through the Confluence desktop app. A system-wide hotkey app doesn't care which. It types wherever the cursor is, so the page editor, the comment box, an inline comment, a live doc, all behave the same whether you're in Chrome or in the standalone app. There's no per-surface setup, no "does it support this view" question. If you can put a cursor in it and type, you can dictate into it.

The same is true for everything else on your screen. The hotkey works in your email, your terminal, a Jira ticket, a Slack message, a Word doc. Confluence is just one of the boxes your cursor visits in a day, and the app treats it like any other. That's the quiet advantage of doing this at the OS level instead of inside one product's editor.

When to just use your OS

If you only need to dictate the occasional comment or a short page, don't install anything. Your computer already does this for free. On Windows, press the Windows logo key + H and voice typing opens in any text box, the Confluence editor included, with no subscription. On a Mac, the built-in Dictation shortcut (or the mic key) does the same, and on Apple silicon it runs on-device with no internet required. For a one-paragraph status update, that's the right tool.

The tradeoffs worth knowing: Windows' Win+H needs an internet connection to run, while macOS dictation and Whisper's local mode don't. And OS dictation gives you the raw transcript, no filler cleanup, no custom vocabulary for your product names. For a quick comment that's fine; you'll edit it anyway.

Pick the smallest tool that solves your problem. For one comment, that's the key you already have. For a 1,500-word runbook you'd otherwise spend an hour typing, the dedicated app, with the AI cleanup turning your spoken draft into something you can almost publish, stops feeling like overkill around the second paragraph you didn't have to type. This pairs naturally with dictating Jira tickets, the other half of an Atlassian day, if writing-by-voice sticks for you.

What Whisper costs

The local dictation pipeline, transcription and the AI cleanup over Ollama, is free for any signed-in user, with no card at signup. So getting your voice into a Confluence page with Whisper costs nothing. Whisper Pro adds the cloud features (OpenAI transcription, cloud AI enhancement, voice web search), and it carries a separate trial. The exact numbers live on the pricing page rather than here, because prices move and a blog post is a bad place to keep them current.

Confluence may ship real dictation one day, the way collaborative editors tend to once enough people ask. Until then, your computer already has the feature for short bits, and a hotkey app exists for the long pages where talking the draft out beats typing it. Two tools, one wall to get over: the typing. Most of the time you need the smaller tool. For the docs nobody wants to write, the bigger one earns its keep.