v1.0.0 - Available Now

Voice at the
speed
of thought.

Stype converts speech to text instantly using offline Whisper AI. No cloud, no subscriptions - your voice stays on your machine.

Windows 10 / 11  ·  4 GB RAM min  ·  No internet required

Live overlay preview
Recording...
100% Offline
<1s Latency
0 Cloud calls
Quickstart

Up and running
in sixty seconds.

Everything you need to go from download to your first dictated sentence.

01
Download & Install

Run the setup executable. Stype launches and parks itself in your Windows system tray. If SmartScreen prompts, click More Info → Run Anyway - expected for independent open-source apps.

02
Bind Your Hotkey

Right-click the tray icon, open Settings, and assign any global shortcut. Default is Ctrl + Space. Works across every app - browser, IDE, email, anywhere.

03
Speak & Paste

Press your hotkey once to begin, speak naturally, press again to finish. Stype shows a live status pill, then pastes the transcribed text directly at your cursor position.

Under the Hood

Engineered for silence,
speed, and precision.

Every decision in Stype is built to stay out of your way while you think.

Absolute Privacy

Zero audio leaves your machine. Whisper AI runs entirely on local hardware. No servers, no logs, no cloud calls.

Near-Zero Latency

faster-whisper with a CTranslate2 C++ backend. High-accuracy transcription on CPU - no GPU required.

Universal Injection

Works in every app - browsers, IDEs, Office. Suppresses OS hotkey conflicts and injects text at your cursor.

Smart Formatting

Automatic capitalization and punctuation. Text arrives clean and paste-ready - no cleanup needed.

Minimal Footprint

Low CPU at idle, modest memory, zero telemetry. Stype is present when you need it, invisible otherwise.

Real-Time Pill UI

A floating overlay cycles through Recording, Processing, and Pasted states. It vanishes when done - no clutter.

Built With
Python 3.13 PySide6 faster-whisper CTranslate2 sounddevice numpy keyboard pyperclip Inno Setup