You think faster than you type.
Typing caps you at 40 to 60 words a minute. You think and speak at roughly three times that. Whisperstream closes the gap without sending a syllable to the cloud.
Works in any app
Speak anywhere you can type.
Works in the apps you already use
Talk like you talk. It writes like you write.
Cleanup with Stream-1 Mini
Say it messy. Get it clean.
Stream-1 Mini, our own cleanup model, cuts filler and applies your spoken corrections without sending a word off your computer.
You said
I need two extra days make that five the discovery phase took longer than planned.
Cleaned up
I need five extra days. The discovery phase took longer than planned.
Per-app styles
Gmail gets an email. Slack gets a message.
Whisperstream sees which app or site you're in and picks the matching style. Write your own prompt for any of them.
You saidhi steven um the launch notes are ready i'll send them thursday no sorry friday morning thanks
Your dictionary
Spells your names right.
Add names, products and jargon once. Cleanup matches them by sound, so a name the speech model heard as "Steven" can still come out as "Stephen".
Agent Speak for Claude Code and Codex
Hear your agent respond. Answer by speaking.
A short summary appears on screen, read aloud if you like. Reply with your usual dictation shortcut, without hunting for the right terminal.
Claude
Fixed the failing import test and updated two files. Ready for you to review.
Speak or type
Privacy
Your voice never leaves your computer.
Whisperstream turns speech into text right on your PC or Mac.
We never receive your recordings or your text, so there is no dictation on our servers to leak or sell.
Stays on your computer
Your voice
Turned into text locally. Never uploaded.
Your text
Typed straight into the app you're using.
Your history
Saved encrypted. Turn it off any time.
The edge of your computer
When it goes online
Downloading a model
At setup
Checking for updates
About once a day
Cloud cleanup
Off until you turn it on
Sends your text, never your audio, to a provider you configure.
Need it stricter? Lockdown Mode turns off all cloud models.
Why rent dictation?Get it free.
Whisperstream
Cloud appsLeading cloud apps
Speech recognition
On your computer
Remote cloud servers
Internet required
Not for dictation
Always
Privacy
No cloud transcription path
Audio sent to cloud servers
Yearly cost
$0
$140+/year
Cost over 3 years
$0
$420+
Account required
No
Yes
Common questions
Start speaking.Completely free.Fully local.
Open source, so you can read every line of code.
Compare Whisperstream
Guides and comparisons
- best dictation software for Windows
See where Whisperstream lands in our full guide.
- Wispr Flow alternatives for Windows
Coming from Wispr Flow? The six best, compared.
- offline dictation software for Windows
No internet connection? Four fully offline picks.
- Whisperstream vs Windows Voice Typing (Win+H)
Already using Win+H? What a dedicated app adds.
- voice input for Claude Code
Dictate prompts instead of typing them.
- how Whisperstream compares with Wispr Flow
Price, privacy, and Windows performance, side by side.