WhooshType Download

Voice typing for macOS and Windows

Say it.ADHD-approved
It’s already typed.

Hold ⌥ and talk the way you think — “um”s, restarts and all. Clean, punctuated text lands in any app, in the right tone.

Download Free to start — see the plans 7 days or 3,000 words, no card

macOS 14+ on Apple Silicon · Windows 10 and 11 · 60+ languages

Mia
Hey! Finished the first draft of the pitch deck and just sent it to you.
Nice, thanks! I’ll go through it tonight 👍
Got a few questions. When can we talk?
Message
Speech recognized ~0.27 s after you let go of the key

4× faster than typing: you speak ~150 words a minute, type ~40

~0.27 s from letting go of the key to recognized text

Two languages in one sentence, terms kept as written

$9.90 a month for Unlimited — or pay as you go with no monthly fee

How it works

Talk messy.
Send it clean.

WhooshType listens while you talk, then reads it over like a sharp copy editor: out go the “um”s and the repeats, in go commas and capitals — and every word that matters stays.

filler words

so like um i think we should uh move the meeting to Thursday, like at three3 o’clock.

changed your mind

send the report on Monday no wait on Tuesday before lunch.

said it twice

the demo is on Thursday, so yeah, Thursday, that’s when we show it. bring the slides, I mean just make sure you have the slides with you.

a dictated list

plan for tomorrow:

first 1. team call

second 2. review pull requests

third 3. write the docs

numbers and dates

we meet on October fifteenth15 at five thirty5:30, budget three hundred thousand dollars$300,000.

terms as written

i pushed the fix to git hubGitHub, but see eyeCI failed on the lint step. we need to fix the es lintESLint config.

And it knows where you’re writing: casual in a chat, polished in an email, word for word in a code editor. Start a list, and the next item gets the next number.

Made for ADHD brains

For minds with
forty tabs open.

A thought shows up in the middle of ten other things, and it won’t wait while you find the right window and type it out. Hold one key and say it as it comes — tangents, restarts and all. It lands as clean text before the next idea pushes it out.

  • No switchingOne key, in whatever app you’re in.
  • No blank pageRamble — it comes out tidy.
  • No lost thoughtsSaid once, written down right where you are.

Modes

Three keys.
Not a single window.

You say

Appears in the field

Plus modes of your own — say, “commit message” or “prompt for an AI”, each with its own key and its own rules.

And also

Little things.
Hard to give up.

  1. Live text

    As you talk, the recognized words run above the pill — you can see you’ve been understood before you finish the sentence.

  2. Fix what you just said

    Tap the key, hold it again and say “make it softer” or “Monday, not Friday” — the text you just pasted is rewritten in place.

  3. A dictionary that learns

    Correct a colleague’s name or a project title once — next time it’s spelled right.

  4. File transcription

    A voice message, a lecture, a meeting recording — pick the file in History and get it as text.

  5. Claude Code by voice

    Dictate tasks to the agent, get a “done” notification, and approve its actions by voice.

Pricing

All you can say.
Or pay as you go.

Two plans, both with every feature. No “credits”, no “tokens”: prices in dollars, tax included.

Unlimited

One flat price, the cloud included.

$9.90a month

Billed monthly. Or $99 a year — 17% less.

$8.25a month

$99 billed once a year — 17% less than monthly.

  • Dictation in the cloud included — and Translate, the Command key and file transcription.
  • Up to 3 devices at once, Mac and Windows in any mix.
  • Cancel anytime in the app.

Pay as you go

No monthly fee — pay only for what you use.

$0a month

Top up a balance in dollars when you need to.

  • Top up from $5 — your balance never expires.
  • The providers’ prices + 20% — for payment processing and taxes.
  • Auto top-up if you want it: from your saved card when the balance drops to $1.

Free to start: 7 days or 3,000 words, whichever comes first. No card. Then pick a plan in the app.

Download

Your words
stay yours.

History and dictionary live on your computer. We don’t keep them on our servers.

Cloud without training. Your voice and text go only to providers that don’t train models on them, and our server keeps none of it.

FAQ

The short version

Which plan should I choose?

Both have every feature. If you dictate every day, Unlimited keeps it simple: one flat price with the cloud included, so there’s no balance to keep an eye on. If you dictate now and then, Pay as you go costs nothing in the months you don’t use it, and what’s left on the balance never expires. Not sure yet? Start free, and switch later in the app.

How is the balance spent?

On Pay as you go, each dictation is speech recognition plus one short pass of a model over the text. You pay what the AI providers charge, plus 20% for payment processing and taxes. The providers’ prices can change; we always charge the current price plus the same 20%. Unused balance never expires. On Unlimited, nothing is taken from the balance.

What happens when the balance runs out?

On Pay as you go, dictation and the Command key pause until you top up. Auto top-up keeps that from happening: when the balance drops to $1, the amount you chose (from $5) is added from your saved card. Or switch to Unlimited in the app.

Can I switch plans?

Yes, in the app, any time. Unlimited starts right away. Moving from Unlimited to Pay as you go keeps Unlimited until the end of the period you’ve paid for; then the balance takes over.

How many devices can I use?

Unlimited covers up to 3 devices at once, Macs and PCs in any mix. With Pay as you go, you can be signed in on up to 10. Signing in on one more asks you which device to sign out — it never signs one out by itself.

Which languages does it understand?

More than 60, including English, Spanish, German, French, Russian and Serbian — and two of them in one sentence, with technical terms kept as written.

Which apps does it work in?

Any text field on Mac and Windows: messengers, mail, the browser, notes, code editors, the terminal. You can set a tone for each app or website.

What can I say to the Command key?

Anything: a question (it can search the web), a message to the person on screen in your style, a translation, an explanation, an edit of the text you selected, or one of your automations. It works out what you meant; if it guessed wrong, say “no, write a message” or “no, just answer”. The result appears in a card: read it, or press Enter to insert it. The window’s text goes with your request, so it knows what you’re looking at, and a screenshot only when the text isn’t enough.

What computer do I need?

A Mac with Apple Silicon on macOS 14 or later, or a PC with Windows 10 or 11 (64-bit).

How is it better than the dictation built into macOS?

Built-in dictation writes down words. WhooshType writes text: it drops filler words and false starts, adds punctuation, makes lists, understands speech that mixes languages and matches the tone to whoever you’re writing to.

Try saying
your next email.

Free for 7 days or 3,000 words, no card · macOS 14+ on Apple Silicon · Windows 10 and 11