cjpais/handy

Handy is a free, open source desktop application that turns spoken words into typed text, running entirely on your own computer. You press a configurable keyboard shortcut, speak, release, and your words appear in whatever text field you have focused, no audio is ever sent to the cloud. It is built for people who want speech-to-text without a subscription and without surrendering their voice to a third-party service. Under the hood, Handy first uses Voice Activity Detection with Silero to filter silence out of the recording, then runs the audio through your choice of local transcription models. You can pick Whisper models in several sizes (Small, Medium, Turbo, Large) with GPU acceleration when your hardware supports it, or Parakeet V3, a CPU-optimised model with automatic language detection. The transcribed text is then pasted directly into the app you were already using. Optional integrations like a Raycast extension let you control recording, browse history, and switch models from outside the app. You would reach for Handy when you want hands-free dictation into emails, code editors, chat apps, or notes, especially when privacy matters or you work offline. The project openly states it is not trying to be the best speech-to-text tool but the most forkable one, so it also appeals to developers who want to extend or rebuild a transcription stack themselves. The application is built with Tauri v2, combining a React and TypeScript frontend styled with Tailwind CSS for the settings interface, and a Rust backend for system integration, audio handling, and machine-learning inference. The full README is longer than what was provided.