Say it. It's typed.

Everywhere your cursor goes.

Nocturne Speak isn't an app you dictate into — it's dictation for every app you already use. Wherever you can type, you can talk: email, documents, code editors, chat, search bars.

One shortcut, and your voice lands right where your cursor is.

Hold a key of your choice — F8 out of the box, or any function key — anywhere, in any app, even with Nocturne tucked away in the tray. Tap it twice and it keeps listening hands-free until you tap again. Or skip the key entirely: set a wake word and say “Nocturne speak…”. For chat boxes and assistant sidebars, it can press Enter for you so the message sends itself.

A small moon card tells you when it’s listening — gold while you dictate — and never steals focus from the window you’re typing into.

No special windows, no copy-paste shuffle — and no audio sent anywhere.

Natural speech, clean text.

Talk the way you actually talk — false starts, "um", thinking out loud. Nocturne Speak keeps up and writes down what you meant: punctuated, capitalized, readable.

It follows the rhythm of real speech instead of forcing you into robot-command cadence.

If a microphone is unplugged, or it hears only silence, it says so plainly. Hands-free listening stops itself after two quiet minutes, and no recording runs longer than thirty.

Your voice is transcribed on your machine — the recording is deleted the moment the words are written, and none of it is ever sent.

Easier on your hands.

A working day is tens of thousands of keystrokes. Speaking is none of them. Dictating your email, notes, and drafts gives your wrists and fingers a genuine rest.

If typing is uncomfortable — long days at the keyboard, sore wrists, hands that have earned a break — your voice can carry most of the writing you do.

Rest your hands without handing your voice to a server.

Think out loud. It comes back structured.

You choose how far it goes. Just the basics transcribes instantly. Polish it takes a second or two — the ums vanish, punctuation lands, dead air disappears. Rewrite it clearly goes further still: ramble for five minutes and the model reshapes it into organized, usable text — an outline, a brief, a spec.

Or leave it on Automatic, and it picks the deepest level your computer can do without making you wait — and tells you which one that is. Spoken lists — “first item, second item” — come out as bullet points at every level.

If you build with AI tools, it's the fastest way to turn a half-formed idea into a prompt worth running. If you write, it's how a story gets spoken into existence.

Five rambling minutes in, one structured page out — and none of it ever leaves your machine.

What all that dictation adds up to.

The Insights tab keeps count — never your words. How many dictations, how many words, how long you spent speaking and how many hours of typing that saved. How fast you talk, how many “ums” slip in, which phrases you lean on, when you dictate most, and which apps your words end up in.

Ask it in plain words — “When do I dictate most?”, “Has my pace changed over the last two weeks?” — and it answers from those counts, in a few sentences, right on your computer. Every number it quotes is one you can see on the tab.

It never reads your transcripts unless you switch that on. When you do, it can also answer about what you actually said, from your most recent transcripts, and it says so under every answer.

Counts only, kept on this computer. Reset clears them and nothing else.

Private by physics, not policy.

Most dictation tools send your voice to a server and send text back. Nocturne Speak has no server to send it to — transcription happens on your processor, in your memory.

That's not a privacy setting you have to trust. It's how the software is built.

Text it pastes is marked so that Windows clipboard history and cloud clipboard sync never keep or upload it. Your transcripts stay in the app only as long as you want: keep them, clear them automatically, or have them expire after thirty minutes, an hour, a day or a week.

Your voice never leaves the room it was spoken in.