Local voice dictation for Windows

Your voice, set in type.
Your words never leave your PC.

Press one hotkey and speak — Stanza writes it at your cursor, in any app on Windows. Powered by Whisper AI running entirely on your machine. No cloud transcription. No audio uploads. No one listening in.

2,000 free words · then $10/month · Windows 10/11

Listening any app · at your cursor

≈1s transcription
0 words uploaded
No word caps
Every Windows app

01 — How it works

Faster than typing, everywhere you type.

1

Press the hotkey

Tap Ctrl + Alt + J in any app — email, docs, code, chat. Stanza wakes instantly.

2

Say it

Speak naturally, as long as you like — no countdown, no caps. Pause to think; Stanza waits. Whisper AI transcribes on your own GPU, entirely offline.

3

It’s written

Your words land at the cursor, already clean — and your clipboard is restored exactly as it was. No window switching, no waiting on a server.

02 — Privacy

Your voice is yours.

Cloud dictation tools stream your audio to someone else’s servers. Stanza’s speech model lives on your machine — your audio and your words stay with you.

Runs 100% local

Whisper AI runs on your own GPU (or CPU). Unplug the internet — Stanza keeps transcribing.

Zero audio uploads

Your voice is transcribed on-device and never uploaded. The single exception is a clip you explicitly choose to share by clicking “report bad transcription.” Your call, case by case.

Your words, your files

Every dictation is saved as plain Markdown on your own disk. Search it, reuse it, delete it — it answers to you.

$ dictated words & audio, uploaded by default 0 — enforced on our servers, not just promised

The exact policy: Stanza sends usage metrics — word counts, durations, app version, device type — to keep the product healthy. Never your text, never your audio. Content leaves your PC only when you click “report bad transcription” on a specific clip, and our backend is built to discard dictation content otherwise, even if it were ever sent.

03 — The dashboard

Know your own voice.

Stanza is the only dictation tool with an editorial analytics dashboard of your speech — beautiful enough to open every morning.

  • Words dictated today, and every day
  • Time saved vs. typing, tracked live
  • Streaks, peak hours, most-spoken words
  • Full history with instant search & audio replay
“A stanza is a unit of verse.
Every dictation you make is one — kept, in your own words.

Everything you dictate is written to plain Markdown files, one per day, on your own disk. Your spoken archive — searchable, reusable, portable. No export button needed, because it was never locked up.

~/Stanza/cache/
2026-07-04.md  — 3,912 words
2026-07-05.md  — 5,208 words
2026-07-06.md  — 4,218 words

04 — The difference

Private and cheaper. Pick both.

Stanza
Cloud dictation apps
Where your audio goes
Nowhere — stays on your PC
Streamed to remote servers
Works offline
Yes, fully
No
Your transcript archive
Markdown files on your disk
Their database
Word limits
None
Capped tiers
Speech analytics
Editorial dashboard
Price
$10/month
$12–15/month

05 — Pricing

One plan. Everything in it.

No tiers, no seat math, no “pro” upsell. Ten dollars a month, and your dictation stays private forever.

Your first 2,000 words are free — try Stanza before a single dollar leaves your pocket.

$10 / month

2,000 free words to start · cancel anytime, one click

  • Unlimited dictation, in any app
  • Whisper AI model included — runs on your PC
  • Editorial analytics dashboard & full history
  • Markdown archive of everything you say
  • English & Portuguese, more coming
  • All updates included
Get Stanza — $10/month

Windows 10/11 · secure checkout by Lemon Squeezy

06 — FAQ

Fair questions.

Does my audio ever leave my computer?

No. Transcription runs entirely on your machine with a local Whisper AI model — unplug the internet and keep dictating. The only way a clip ever leaves your PC is if you explicitly click “report bad transcription” to share that one case with us.

What data does Stanza collect?

Usage metrics only: word counts, dictation durations, app version, device type. Never your text, never your audio. And it’s structural, not a pinky promise — our servers discard dictation content by design, so normal dictations couldn’t be stored even if they were sent. Full details in our Privacy Policy.

What do I need to run Stanza?

Windows 10 or 11. An NVIDIA GPU makes transcription feel instant; without one, Stanza falls back to CPU — still private, a touch slower.

Does it work in every app?

Anywhere you can place a cursor: email, browsers, Word, Notion, Slack, your IDE. Stanza types at the cursor like a very fast keyboard.

Which languages are supported?

English and Portuguese today, with more Whisper-supported languages on the way.

Is there a free trial?

Yes — your first 2,000 dictated words are free. No credit card, no clock ticking, no weekly reset games: take a day or a month to use them. When you hit the limit, it’s $10/month to keep dictating.

How is Stanza different from Wispr Flow?

Wispr Flow transcribes in the cloud — your audio goes to their servers, and your history lives in their account. Stanza transcribes on your device, keeps your history in files you own, adds a speech-analytics dashboard no other tool has, and costs $10 instead of $15 a month.

Can I cancel anytime?

Anytime, one click, no email needed: open your billing portal (linked in your purchase receipt and in the site footer under “Manage subscription”) and hit cancel. Your Markdown archive stays on your disk — it was always yours.

Your best writing
was spoken.

Dictate at the speed of thought — and keep every word to yourself.

Get Stanza — $10/month