Dictation for writers
Speaking is roughly three times faster than typing. Here is how writers use VocalCraft to get a messy first draft down fast — and why the editing stays manual.
Dictate with two keys
Tap Caps Lock to arm VocalCraft — the LED turns on. Click any text box, tap Spacebar once to start listening, speak naturally, tap Spacebar again. Your words paste straight into the field you clicked. That's the whole workflow.
The writer's bottleneck is the first draft
Typing caps most people around 40–60 words per minute. Speaking runs 120–180. Every writer knows the feeling: the scene is clear in your head, but your fingers can't keep up, and by the time the sentence is typed the next two have evaporated. Dictation fixes the throughput problem — you talk the draft, then you edit with the keyboard, which is where the craft actually happens anyway.
A dictation workflow that actually works
- Arm with Caps Lock, dictate with Spacebar. No app to open, no recording button. The microphone lives on a key you already touch.
- Speak punctuation as you go. Whisper handles natural phrasing; say "comma", "new paragraph" when you want structure.
- Draft in whatever you're already using. Word, Google Docs, Scrivener, Obsidian, Notion in the browser — VocalCraft pastes into the focused field in any of them.
- Never lose a draft to the cloud. Your manuscript is processed on your machine. Unpublished work stays unpublished.
Why not just use Windows Voice Typing or a cloud app?
Windows Voice Typing works, but it's tethered to Microsoft's cloud and likes to stop listening at awkward moments. Cloud dictation apps send every word — including your unpublished chapters and your client's name — to a server you don't control. VocalCraft runs a full neural network locally: no round-trip lag, no quota, and a draft that exists only on your disk. If you write anything confidential, that difference matters.
Prefer a head-to-head? See VocalCraft vs Dragon NaturallySpeaking and VocalCraft vs Windows Voice Typing.
Frequently asked questions
Is VocalCraft good for writing a book?
It is excellent for first drafts. Most authors dictate the raw chapter, then edit on keyboard. The speed win is on the draft; the craft stays in the edit.
Does it handle Australian/British spelling and accents?
The Whisper models are robust across accents, and because processing is local you can pick the model that suits your voice. You can also just say the word and fix spelling in the edit pass, same as typing.
What if I dictate something private?
Nothing is uploaded — there is no upload path. The audio is processed in memory on your PC and the text is pasted locally. For journalists, ghostwriters and anyone under NDA, that is the whole point.