Voca · voice to text in your browser, or on Windows, macOS, and Linux

Dictate anywhere.
Your words, nobody else's.

Speak, and Voca turns it into text right in your browser, no install needed. Or grab the desktop app for Windows, macOS, or Linux to dictate anywhere with a system-wide hotkey. Audio is deleted the moment it's transcribed. We never keep it, train on it, or sell it.

No install, no card · works in any modern browser
0secondsof audio transcribed
Recording
00:00:00
""
01 · Privacy

Audio deleted instantly.

Your audio takes a one-way trip: your desktop, our server, transcription, then deleted. We don't keep it, back it up, or analyse it.

02 · Pricing
$6 / ∞

Pay for what you say.

$6 a month, unlimited. Or top up prepaid credits at ten cents an hour and never think about a subscription. No surprise charges, no silent caps, no cancel-and-keep-charging tricks.

03 · Speed
0.4s

Speech to text, fast.

Built on the quickest open inference stack in production. Finish a sentence and the text is already there. No buffering wheel, no "processing…" spinner, no waiting to paste.

Three steps.

01 · CAPTURE

Press the hotkey.

The desktop app starts recording locally. No browser tab, no always-on mic, no background listening.

CtrlShift;
02 · TRANSCRIBE

Transcribed, then deleted.

Your audio is sent to our server, forwarded to transcription, and deleted the instant the text comes back. No backups, no logs of what you said. Just a duration and a timestamp for billing.

03 · RETURN

Text returns

The transcript is returned to your computer and presented in front of you.

"Worth asking what we'd lose if we cut everything else."
0
seconds of audio retained
$6
flat · unlimited pro
30min
free every month
0.4s
median latency
Privacy

No backups.
No training.
No logs.

Most dictation tools keep your audio around: to "improve quality," to train next quarter's model, or because nobody thought hard about deleting it. Voca was built the other way around. Your audio exists on our server only long enough to be transcribed, then it's gone.

  • Audio deleted immediately after transcription. Not backed up, not queued, not analysed.
  • Transcripts stored only if you opt in. Off by default, deletable anytime.
  • No models trained on your voice or your text. Full stop.
  • Metadata kept for billing: duration, timestamp, model. Nothing more.
01 · Client
Browser or desktop
Captures audio in your browser, or locally on desktop when you hold the hotkey.
02 · Server
Voca API
Forwards audio to transcription. Deleted as soon as text returns.
03 · Return
Transcript
Text comes back to your clipboard. That's the whole trip.
Pricing

Pay for what you say.
Nothing more.

Thirty free minutes every month. No card required.

Free
$0/forever
  • 30 minutes of transcription a month
  • Browser & desktop app, Turbo model
  • No credit card required
Start free
Pay as you go
$0.10/hour
  • Top up from $2 (≈ 20 hrs audio)
  • Credits never expire
  • Use across devices on your account
Buy credits

Speak.
Done.

Try it in your browser in seconds, no install, no card. Or take the desktop app anywhere on Windows, macOS, or Linux.