100% LOCAL • NO CLOUD • NO TELEMETRY

Your voice,
your machine,
your text.

VoiceInk transcribes your speech locally on your Windows PC. Hold a hotkey, speak, release — your words appear in any app. No internet, no cloud, no compromises.

Inspired by WisprType • Built for Windows • Open Source (MIT)

LISTENING
READY
99+ Languages
CPU + CUDA GPU
Zero Data Leaves PC
MIT License
Pipeline

Hold. Speak. Done.

From hotkey press to text in your app — the entire pipeline runs on your machine in under a second.

01

Hold Hotkey

Press and hold your configured hotkey (default: Right Ctrl). Recording starts instantly from your default mic.

Push-to-talk
02

Speak Naturally

Talk in any of 99+ languages. Auto-detection handles the rest. Audio stays in memory — never written to disk.

16kHz mono PCM
03

Transcribe Locally

faster-whisper runs entirely on your GPU or CPU. No data sent anywhere. ~0.3s on CUDA, ~2-5s on CPU.

CUDA or CPU
04

Text Appears

Transcribed text is injected into whatever app you're using — Word, Slack, VS Code, browsers, anywhere with a cursor.

Win32 SendInput
voiceink.exe
Hotkey pressed — recording started
Hotkey released — recording stopped (2.4s)
Transcribing with faster-whisper (small, CUDA)...
Transcription complete (0.3s)
"Hey team, I've pushed the latest changes to the PyTorch branch. Can someone review the transformer module before we merge?"
Text injected into VS Code
Features

Everything you need,
nothing you don't

Every feature is designed for speed, privacy, and reliability. No bloat, no subscriptions, no cloud dependency.

Smart Typing

Automatically removes filler words (um, uh, like), fixes self-corrections, adds punctuation, and capitalizes sentences.

Filler Removal Auto Punctuation Offline

Custom Dictionary

Add names, technical terms, medical vocabulary, and brand names. Injected into Whisper's prompt for better accuracy.

Names Technical Medical

Global Hotkey

System-wide push-to-talk works in any app. Configurable — Right Ctrl, Right Alt, CapsLock, F13, or custom combos.

Push-to-talk Customizable

Works Everywhere

Text appears in any app with a text cursor — Word, Outlook, VS Code, Slack, Discord, Chrome, Notion, and more.

Win32 API Clipboard Fallback

System Tray App

Lives quietly in your taskbar. Icon changes color to show state — idle (gray), recording (red), processing (yellow), done (green).

Idle
Rec
Proc
Done

History & Search

Every transcription saved locally in SQLite. Search, copy, delete, or export as TXT/CSV. Never leaves your device.

SQLite Export Search

Works in any app with a text cursor

Microsoft Word Outlook VS Code Chrome Slack Discord Notion ChatGPT Gmail Notepad Firefox Terminal
Microsoft Word Outlook VS Code Chrome Slack Discord Notion ChatGPT Gmail Notepad Firefox Terminal
Privacy First

Your voice never
leaves your machine

VoiceInk is built with a zero-trust network model. It literally doesn't need the internet after the initial model download. No telemetry, no analytics, no accounts, no cloud — just you and your PC.

Works Offline
No Telemetry

No Internet Required

After model download, works completely offline. No API calls, no web requests.

No Telemetry

Zero analytics, zero tracking, zero phone-home. We don't even collect crash reports.

Audio Stays Local

Audio is processed in-memory and discarded. Never written to disk, never uploaded.

Local Storage Only

Transcription history stored in local SQLite. Never synced, never uploaded, fully yours.

No Account Needed

No sign-up, no login, no email. Download and run. That's it.

Open Source (MIT)

Every line of code is auditable. Verify our claims yourself. Fork it, modify it, share it.

Whisper Models

Pick your accuracy vs speed

Five model sizes for every use case. Smaller models are faster; larger models are more accurate. All run locally.

tiny
Size75 MB
SpeedFastest
AccuracyBasic

Quick notes

base
Size145 MB
SpeedFast
AccuracyGood

General use

Recommended
small
Size466 MB
SpeedModerate
AccuracyBetter

Best balance

medium
Size1.5 GB
SpeedSlow
AccuracyHigh

Professional use

large-v3
Size3 GB
SpeedSlowest
AccuracyBest

Maximum accuracy

NVIDIA CUDA

~0.3s for 10s audio • float16 • Requires CUDA toolkit

CPU Fallback

~2-5s for 10s audio • int8 • Works on any PC

Comparison

VoiceInk vs WisprType

WisprType brought local voice-to-text to macOS. VoiceInk brings it to Windows — and makes it open source.

Feature WisprType macOS VoiceInk Windows
Platform macOS (Apple Silicon) Windows 10/11
Local Processing
Smart Typing
Custom Dictionary
Open Source
GPU Support Apple Neural Engine NVIDIA CUDA + CPU
Cost Free Free (MIT License)
Windows Support
Linux Support Planned
Languages 99+ 99+
Architecture

Built with modular precision

Clean, auditable Python architecture. Each module has one job. Easy to extend, easy to audit.

audio_capture.py

Records 16kHz mono PCM via sounddevice. In-memory numpy array. No disk writes.

whisper_engine.py

faster-whisper wrapper. CUDA/CPU auto-detect. Custom vocab via initial_prompt.

smart_typing.py

Regex-based filler removal + punctuation. Optional LLM post-processing.

text_injector.py

Win32 SendInput or clipboard+paste. Restores original clipboard content.

hotkey_manager.py

Global push-to-talk via keyboard library. Non-blocking background thread.

custom_dictionary.py

JSON-based vocab manager. Categories: Names, Technical, Medical, Legal, Brands.

ui/tray.py

pystray system tray with state icons. Menu for quick toggles and settings.

history.py

SQLite local storage. Search, copy, delete, export TXT/CSV. Never uploaded.

Roadmap

From prototype to product

A phased approach to building a production-ready, distributable Windows app.

Phase 1 1-2 days

MVP Prototype

Working proof-of-concept with core pipeline.

Hold hotkey → record faster-whisper CPU Auto-paste
Phase 2 1-2 weeks

Daily Driver

Polished enough for daily personal use.

System tray Settings JSON Smart Typing Custom dictionary Model selection
Phase 3 1-2 weeks

Full GUI App

Complete graphical interface for all settings.

PyQt6 settings History window Dictionary editor Audio visualizer
Phase 4 1 week

Distribution

Packaged installer for non-technical users.

PyInstaller .exe Inno Setup installer Auto-updater Startup entry

Future Roadmap

AMD GPU support via DirectML
macOS port
Real-time streaming transcription
Voice commands ("new paragraph")
Per-app tone adaptation
Plugin system for post-processors

Ready to type with your voice?

Download VoiceInk, pick your Whisper model, and start talking. No account needed. No internet required after setup. Just your voice, your machine, your text.

Windows 10/11 NVIDIA CUDA + CPU MIT License ~470 MB (small model)

Or build from source:

$ git clone https://github.com/voiceink/voiceink
$ cd voiceink && pip install -r requirements.txt
$ python main.py