Powered by Rust & Whisper AI

Type Everywhere with Your Voice.

An ultra-minimal, universal voice-to-text overlay engine. Completely local, private by default, and active across all apps with a single keystroke.

100% Offline Privacy
~142MB Model Size
< 15ms OS Inject Latency
gemini-cli — VoiceTyping Session
# Press Ctrl+Alt+/ or click the widget tray to toggle input
$ gemini prompt --stream --local
[System]: Initializing Whisper AI Engine context link...
[System]: Listening for user vocal overlay streaming...

|
⋮⋮
Click to Record / Ctrl+Alt+/
Status: Idle

Interactive Dual-Engine Architecture

Choose complete offline data security or premium cloud scaling parameters based on your work environment.

Feature Metric Local Engine (Free Tier) Cloud Engine (Premium Tier)
Privacy Footprint 100% Offline

Zero telemetry. Audio is processed directly in local memory and instantly discarded.

Secure Cloud Routing

TLS 1.3 encrypted streaming. Zero caching policy on cloud servers.

Supported Languages English Optimized

Includes custom vocabularies and shorthand expansions for coders.

99+ Languages

Seamless multilingual dictation with auto-language detection.

On-Device Profile Whisper CPP Binary

Uses ~142MB quantized Whisper model. Runs efficiently on CPU/GPU.

Zero Overhead

No disk footprint or model file downloads required. Ideal for older hardware.

Operational Cost Free Forever

Download the application and run indefinitely without a billing account.

SaaS Subscription

Authenticated instantly using Magic Link verification codes.

Download VoiceTyping

Select the pre-compiled binary package matching your workstation architecture.

🍎

macOS

Apple Silicon & Intel Binaries

Requires macOS 10.15+. Optimized for both ARM64 and x64 processor architectures.

Apple Silicon (M1/M2/M3)
Intel (x64)
🪟

Windows

ARM & x64 Installer Packages

Supports Windows 10/11. Optimized for ARM64 and standard x64 architectures.

64-bit (x64)
ARM64
32-bit (x86)
🐧

Linux

x86_64 / x64 Linux Packages

Supports major Linux distributions (Ubuntu, Debian, Fedora, Arch). Native pre-compiled binaries for x86_64.

64-bit (x86_64)
ARM64
Coming Soon

Frequently Answered Capabilities

How does Voice Typing input text into third-party apps like Slack, Word, or VS Code?

Voice Typing utilizes an internal OS background injection script. As soon as the active local or cloud audio pipeline translates voice vectors to a text string, the software's Rust layer simulates hardware input events, effectively typing characters directly into whatever input element holds focus on your desktop.

Does Voice Typing require cloud subscription models or cloud API connections?

It depends on your required tier. The free tier of Voice Typing features a completely offline engine core running an isolated, highly optimized local instance of OpenAI's Whisper model for English language dictation with zero network dependencies. The premium tier utilizes an online, cloud-based network model to process advanced speech matrices.

Is Voice Typing free or commercial, and what languages are supported?

Voice Typing features a hybrid model handled entirely via secure Paddle payment processors. The free tier is strictly offline and exclusively supports the English language. Upgrading to the premium commercial tier unlocks an online cloud-based system featuring support for over 99 languages.