Free voice typing for writing and coding. Speak as you think — directly into any editor, browser, and messenger.
Voice typing built directly into the system: no extra windows, clipboard hops, or context switching.
Press Alt + V in any active application. While held down, Aura records voice input smoothly in the background.
Whisper and Parakeet models process audio directly on your PC — powered by CPU or NVIDIA GPU.
Formatted text with automatic punctuation and capitalization appears directly where your cursor is positioned.
Engineered for maximum performance, data privacy, and instant response.
When choosing local models, audio never leaves your PC. Zero external cloud servers, hidden logging, or telemetry.
Transcribed text is typed directly under your cursor into any active Windows window without clipboard interference.
Telegram
Discord
VS Code
Notion
Obsidian
Google Chrome
Runs on any Intel and AMD processor out of the box, with 1-click NVIDIA CUDA GPU acceleration.
/
Completely free, with zero ads or paid subscriptions. Source code is open to everyone on GitHub under AGPL-3.0.
Clean and focused settings dashboard. Full control over models, audio capture, and hotkeys.
Launch the app automatically when starting Windows.
Select the language for settings and application notifications.
Automatically pauses hotkey hooks and unloads models from video memory for maximum FPS in games.
Manual toggle is also available via the Aura system tray menu.
Releases video memory (VRAM) when voice typing is unused for the selected duration.
Aura contacts GitHub only when you check manually or enable automatic checks.
Flexible architecture supports both local neural networks and high-performance cloud APIs.
Model stays resident in RAM and transcribes without network lag.
From lightweight Base for any PC to high-accuracy Large v3 Turbo.
Automatic punctuation, capitalization, and formatting.
0.6B model optimized for NVIDIA discrete graphics.
Ultra-low latency streaming voice input without perceptible lag.
Acceleration components download directly inside app settings.
Support for any OpenAI-compatible API endpoint.
Dictate smoothly even on lightweight laptops without dedicated GPU.
API keys are securely encrypted by native Windows DPAPI.
How Aura compares to proprietary cloud tools and built-in Windows voice typing.
| Parameter | Aura | Cloud Services | Windows Voice (Win + H) |
|---|---|---|---|
| Response Speed | Zero network latency (local) | Network audio upload delay | Noticeable latency |
| Privacy | 100% Local on your PC | Audio sent to third-party cloud | Telemetry collected |
| Punctuation | Automatic | Basic | Voice commands only |
| Pricing | 100% Free | Paid subscription | Free |
| Offline Mode | Autonomous (local models) | Requires internet | Requires internet |
