Voice to text in Logseq

Does Logseq have built-in dictation?

No. The Logseq desktop app has no built-in speech-to-text, dictation, or voice-typing feature for writing into a block by voice. There is no microphone button on a block, no voice command, no hidden preference. If you've been combing Settings for it, you can stop. It isn't there.

What does exist is a set of community plugins with "whisper" in the name, and this is where people get turned around. Those plugins transcribe an audio file or a YouTube link into text after the fact. They are useful, but they are not live dictation. You can't put your cursor in today's journal, talk, and watch words appear. They process a recording; they don't type for you while you think. Conflating the two costs an afternoon, and I'd rather you skip that afternoon.

The mobile picture is its own thing, and worth one sentence so you don't chase it on the wrong device: the newer Logseq mobile app has added some voice transcription, but that's a phone feature, and on a phone you'd just use the keyboard's microphone anyway. On the desktop graph most people actually live in, you need a tool that sits on top of Logseq. There are three honest categories, and the rest of this guide covers them.

Press a hotkey, talk, text lands in the block

This is the whole mechanic, and it's boring in the best way. You press a hotkey, you speak, you release, and the transcript pastes at your cursor, in whatever text field has focus. Whisper holds a short tail after you let go of the key, so your last word doesn't get clipped. Because it pastes at the OS cursor, a Logseq block is just "any text box." Desktop app or the browser version, same behaviour.

That's the part the landing pages overcomplicate. There's no plugin to install into Logseq, no API token to paste, no sync job to babysit. Your cursor is in a block, you talk, the words appear in the block. A small capsule shows up while you speak so you know it's listening:

The hotkey is the one thing worth getting right up front. On Windows it's Ctrl+Space; on Mac it's Command+Option, a modifier-only push-to-talk you hold while speaking. Both are changeable in Settings if they clash with something you already use. (My younger daughter once told me a hotkey "didn't work" in her drawing app. It was a conflict, not a bug, which is how I learned the average person has no idea what a hotkey conflict even is. So now every hotkey is customisable.) If you've ever set up dictation on Windows or on Mac, this is the same muscle memory pointed at a different app.

Set it up in two minutes (Windows or Mac)

You need a Mac on Apple Silicon or a Windows 10-or-newer PC, a working microphone, and Logseq open in either the desktop app or the browser. The whole local pipeline is free for any signed-in account, with no payment method asked for at sign-up. Here's the sequence.

The slow part is the model download, not the setup. Everything else is the four steps above. Once it's running, the act of capturing a thought into your graph stops being a typing task and starts being a talking task.

voice to text on Windows · on Mac

A Logseq plugin vs. a system-wide hotkey

Most pages ranking for this keyword point you at a Logseq plugin or at Blurt, a dedicated Mac menu-bar tool that speaks straight into your outline. Those are fine answers, with one structural catch each. The whisper-style plugins transcribe audio files, not live speech into the block you're editing. And Blurt, by its own description, is macOS only — if you're on Windows, it isn't an option at all.

A system-wide hotkey sidesteps both limits. It pastes at the OS cursor regardless of which window owns it, so the same key that fills a Logseq block also fills your Gmail compose box, a Slack message, and a commit message. One tool, every text field, on both Windows and Mac. You don't relearn anything when you switch apps, and you don't need a different solution depending on your laptop.

If you're on a Mac and you only ever capture inside Logseq, Blurt is a tidy, focused pick and worth a look. The moment you're on Windows, or you want the same flow across every program you open, the system-wide route wins. I'd reach for the one hotkey because I switch apps roughly forty times an hour and don't want forty different dictation buttons to remember.

Local or cloud: which mode for a private graph

For Logseq, try local mode first. The whole reason a lot of us chose Logseq is that it's local-first plain text — a meeting recap, a half-formed idea, a journal entry you'd never want on someone's server. It would be a strange choice to keep your notes on your own disk and then route your voice through a cloud to get there. If your Mac is Apple Silicon or your PC is from the last few years, local handles everyday dictation without complaint, and cloud becomes the escape hatch rather than the default.

Here's how the three paths differ, because the app makes you pick and I'd rather you pick well:

• Local Parakeet — NVIDIA's TDT engine, around 600 MB, and the fastest local option — 5 to 10 times faster than Whisper on CPU. Covers English plus 24 other European languages, 25 in total. No translate-to-English. If you journal in English or another European language, this is the quick, fully offline pick.
• Local Whisper — slower than Parakeet on the same machine, but the multilingual builds cover 99 languages and can translate to English. The English-only builds are English-only, not 99. Pick this for Chinese, Japanese, Korean, or any translation work, which Parakeet can't do. Default English model is around 480 MB.
• Cloud (OpenAI, BYOK) — best accuracy and web access, using your own OpenAI key billed straight by OpenAI. Transcription runs on gpt-4o-mini-transcribe by default. Needs internet, so it's the one path that leaves your machine. The Cloud surface is part of Whisper Pro.

The boring truth is that for the kind of text most people put in Logseq, local is plenty. Both local engines run fully on your machine with nothing sent to a server, which is the entire point of a local-first graph. Cloud earns its place when you want top-tier accuracy on a hard recording or you need the model to pull a fact off the web mid-sentence. For a daily-journal habit, start local and only reach for cloud when local leaves you wanting.

Punctuation, blocks, and Logseq markdown by voice

Raw dictation comes out as a run-on. You say "okay so review the architecture doc tag it project alpha and remind me Thursday," and that's the unpunctuated wall any speech engine hands you. Cleaning it up is where the paths diverge.

Windows Voice Typing adds punctuation as you speak, and macOS Dictation handles basic punctuation when you say "comma" or "period." For heavier cleanup — stripping the "ums," fixing the run-ons, turning a spoken paragraph into something you'd actually keep in your graph — Whisper can run an AI pass. Say the activation phrase "Hey whisper" and the text gets enhanced before it lands. On a local model that runs through Ollama; in cloud mode it's gpt-5-mini by default.

For Logseq's own structure — nested blocks, the #tag and [[page]] links, TODO markers — the honest answer is that voice gets you the text and Logseq's own syntax gets you the structure. Dictate the sentence, then type the Tab to indent, the # for a tag, or the [[ for a page link the way you always do. No dictation tool conjures Logseq's outline syntax into existence on command; anyone promising "say tag project alpha and watch it link" is selling you a demo, not a Tuesday. Get the words down fast by voice, shape the blocks with the keys you already know.

That same speak-then-clean flow pays off well beyond your graph — you can also dictate clean prose into any app with the one hotkey, so a long block becomes a few spoken sentences instead of a paragraph you type out.

When to skip a dictation tool for Logseq

Sometimes the right tool is the free one already on your machine, and pretending otherwise would be dishonest. If you only drop short captures into Logseq — a quick journal line, a two-word reminder — your operating system covers it for nothing.

On Windows, press Windows key + H and the built-in Voice Typing bar opens wherever your cursor is, a Logseq block included. It punctuates on its own and is fine for short bursts. The catch: it routes through Microsoft's servers and needs an internet connection, so it isn't an offline option, which matters more than usual when the whole point of your graph is staying local. On Mac, Dictation lets you speak to enter text anywhere you can type, set up in System Settings under Keyboard, and on Apple Silicon general text can be processed on-device. And if you're a Mac user who lives entirely inside Logseq, Blurt is a focused, native pick built for exactly that.

Reach for a dedicated, system-wide tool when the built-ins start hurting: long notes, multilingual work, offline privacy on Windows, or wanting one hotkey that behaves the same in Logseq, your email, and your editor. Below that bar, use what's free. I'm not going to tell you to install an app for a one-line reminder.

The same trade-off shows up if you also keep notes elsewhere — the logic in dictating into Obsidian is identical, because both are local-first markdown apps where the cursor, not a plugin, is the real integration.

Logseq never shipped a microphone button, and after writing this I'm fairly sure it never will. It doesn't need to, because the cursor is the integration. Talk into the block, get text, shape it with the syntax you already know. I dictated most of this guide into a text box that wasn't Logseq, with a tool that doesn't care which box it is, then pasted the lot into my own graph. That's the whole trick.