Features
Offline speech-to-text that works in every app.
WhisperVoca is an offline dictation app for Mac and Windows. Hold a shortcut, talk, and your words appear wherever you are writing. Your voice is transcribed locally by Whisper AI, so nothing is ever sent to the cloud. No account, no subscription, no tracking.
100% local · No account required · Mac & Windows · Powered by Whisper AI
- 100+
- Languages supported
- 100%
- On-device, works offline
- 0
- Audio uploaded or stored
- 8
- Dictation modes
GPU-accelerated on Mac with Apple Metal. No account, no mandatory subscription, free to try.
Why it saves you time
Talk at the speed you think.
Speaking is far faster than typing for almost everyone. Hold a shortcut, say the sentence, and it is on the page, so the emails, messages, notes, and first drafts that used to take minutes take seconds. Across a week of writing, that is hours handed back to you, without sending a second of audio to the cloud.
up to 5x faster than typing
Typical typing runs around 45 words a minute; speaking is far quicker. How much you save depends on the model you pick and how you write.
The basics, done right
Fast, private, and out of your way.
Hold to record, release to paste
Hold your global hotkey in any app, speak naturally, then let go. Your words appear instantly in whatever text field is focused: email, docs, chat, code editors, search bars. There is no separate window to open and no copy-paste step.
Runs fully offline
Download a speech model once and dictate anywhere after that, whether on a plane, in a basement, or with Wi-Fi switched off. Offline speech to text means there is no round trip to a server, so transcription starts the moment you start talking.
Your voice never leaves your device
Audio is transcribed locally by Whisper AI and discarded immediately. Nothing is uploaded, no recordings are stored, and there is no account to create. Privacy is the default, not a paid add-on.
Why offline dictation matters
Most voice typing tools record your microphone and send the audio to a remote server to be transcribed. That means your words, and your voice, leave your computer every time you dictate. It also means dictation stops working the moment your connection drops.
WhisperVoca takes the opposite approach. The speech model lives on your machine, so transcription happens entirely on-device. There is no upload, no server, and no waiting on a network round trip. You can dictate on a flight, in a quiet office, or anywhere your connection is unreliable, and the experience is exactly the same.
Because nothing is uploaded, there is also nothing to leak. Audio is processed and immediately discarded, and the only thing kept is the text history you choose to keep. For anyone handling sensitive writing, such as medical notes, legal drafts, financial details, or private journaling, that difference is the whole point.
Everything it does
Built for real, everyday writing.
Works in any app
WhisperVoca pastes at the operating-system level, so it works in any program with a text field: Slack, Word, Notion, VS Code, Gmail, your browser's address bar, and everything else.
8 dictation modes
General, Email, Slack, Code, Dictation, Medical, Legal, and Finance modes adapt formatting and vocabulary to what you are writing, so the output reads the way that context expects.
Filler-word removal
"Um," "uh," and "you know" are stripped automatically. You get clean, polished text without editing afterward.
Hallucination filter
Silence and empty recordings are discarded, so Whisper never pastes stray phrases like "thank you for watching" into your document.
Custom vocabulary
Teach WhisperVoca the names, brands, and jargon you use most. It can also learn from the corrections you make in your history.
100+ languages
Dictate in over a hundred languages using the same local Whisper models, with no extra downloads per language beyond the model itself.
Transcription history
A local log, grouped by day, keeps your past dictations one click away. It stays on your machine, and only if you choose to keep it.
Multiple speech models
Choose from Tiny to Small models depending on the balance you want between speed and accuracy. Download, switch, or delete them from inside the app.
GPU-accelerated on Mac
Apple Metal powers near-instant transcription on Mac, giving you fast results without a cloud subscription doing the work for you.
Works where you code
Dictate prompts straight into Cursor, Claude Code, Codex, and Amp. Hold the hotkey, talk through the change, and your words land in whichever editor is focused.
Customizable overlay
Recolor the pill and waveform, drag the overlay anywhere on screen, or hide it for an hour when you need the space. It stays out of your way.
Microphone picker
Use your built-in mic, a headset, or a USB mic, or simply follow the system default. Switch inputs without leaving what you are doing.
Quick controls
A native right-click menu on the overlay puts settings, mic selection, hide options, and quit one click away.
Pay once, no subscription
There is a free tier, and paid plans are a one-time or yearly purchase. No monthly fees, no feature paywalls, and no account to create.
Who it's for
Made for people who write a lot.
Writers and students
Draft essays, articles, and notes at the speed you think instead of the speed you type.
Developers
Dictate commit messages, comments, and documentation without leaving your editor, using Code mode for clean formatting.
Anyone who values privacy
If sending your voice to a cloud service is a non-starter, local-only dictation is built for you.
Curious how it compares to cloud-based tools, or what it costs? See WhisperVoca pricing.
Frequently asked questions
Does WhisperVoca really work without an internet connection?
Yes. Once you download a speech model, all transcription happens on your device. You can dictate with Wi-Fi turned off and the experience is identical.
Is my voice sent to the cloud?
No. Audio is processed locally by Whisper AI and discarded right after it is transcribed. Nothing is uploaded to a server, and no recordings are stored.
Which apps does it work in?
Any app with a text field. WhisperVoca pastes at the operating-system level, so it works in email clients, documents, chat apps, code editors, browsers, and more.
Do I need an account or a subscription?
No account is required, and there is no mandatory subscription. There is a free tier, and paid plans are a one-time or yearly option. You can see the details on the pricing page.
Which platforms are supported?
WhisperVoca runs on both macOS and Windows, with the same features on each. On Mac, transcription is GPU-accelerated using Apple Metal.