FeaturesPrivacyCompareFAQDownload

Private dictation
for every app.

Hold your hotkey, speak, release.
Fing pastes the result into any app — fully locally.

macOS & WindowsNo accountNo app telemetry

Hold. Speak. Paste.

A global shortcut turns your speech into text wherever you can type.

Powered by Whisper
GPU-accelerated (Metal / Vulkan)
Built with Tauri + Rust

Local dictation without
the usual tradeoffs.

Fing stays quiet in your tray until you need it, then gives you fast transcription, custom vocabulary, and searchable local history without sending audio to the cloud.

100% local processing

No servers, no third-party APIs, no app telemetry. Whisper runs entirely on-device and the mic is active only while you hold the hotkey.Runs on-device. Mic on only while you hold the hotkey.

GPU-accelerated speed

Optimized for low-latency transcription on Metal (macOS) and Vulkan (Windows), using your GPU to keep inference fast and responsive.Low-latency on Metal (macOS) and Vulkan (Windows).

Multi-language

English, German, Spanish, and French supported out of the box, with automatic language detection so you never have to switch settings.EN, DE, ES, FR with auto-detection.

Custom hotkey

Bind any key or modifier combination to your dictation trigger. Works globally across every app on your system.Bind any key. Works in every app.

Personal dictionary

Teach Fing names, jargon, product terms, and abbreviations so dictation sounds like you — not like a generic voice model.Teach it names, jargon, and product terms.

Lives in your system tray

No persistent window. No dock clutter. Fing sits quietly in your menu bar until you need it, then gets out of the way.No window, no dock clutter — just the menu bar.
And more:Optional local historyAuto-starts with your systemMIT licensed, free forever

Zero data.
Zero cloud.
Zero compromise.

Your microphone is active only while you hold the hotkey. Audio, transcripts, and history stay on your device unless you paste or copy them somewhere yourself.

Data flow
Your voice
stays local
Whisper model
on-device
Transcribed text
your clipboard
blocked
Cloud server
never
Third-party API
never
App analytics / telemetry
never

Dictation you control.

Cloud tools can mean accounts, subscriptions, and unclear audio handling. Fing keeps the full transcription loop local, free, and inspectable.

Feature comparison between Fing and paid or cloud speech-to-text alternatives
FeatureFingPaid / cloud alternativesOthers
PriceFreeSubscription or usage-basedPaid
Runs locallyLocalAlwaysVaries by product and plan~
OfflineYesLimited or unavailable
No account requiredNo accountYesUsually no
Audio privacyPrivacyNever leaves deviceDepends on provider~
Open sourceOpen sourceMITProprietary
WindowsYesVaries~
Custom dictionaryDictionaryYesOften~
GPU accelerationGPUMetal & VulkanVaries~

Common questions.

Does Fing send audio to the cloud?

No. Whisper runs on-device and Fing does not use app telemetry.

Does it work offline?

Yes. After the initial model download, transcription runs entirely locally with no internet required.

Where does the text go?

Fing pastes it into the active app. It can also keep an optional local history that auto-clears after 30 days.

Is it free?

Yes. Fing is MIT licensed and open source — free to use, inspect, and modify.

Why does Windows show a security warning?

SmartScreen may flag new open-source releases from unrecognized publishers. The download is published through GitHub Releases and can be reviewed in the open-source repository.

Download private dictation
in under a minute.

Fing is free, open source, and runs locally on macOS and Windows. No account. No card. No cloud.

macOS 11+ & Windows 10+