A native macOS application that transcribes speech to text locally using AI for private voice dictation.

VoiceInk is an open-source macOS application designed to transcribe spoken audio into text in real time across all installed applications on the system. The software serves as a productivity tool for users who prefer voice dictation over typing for drafting documents, emails, or messages.
The software operates as a desktop utility that integrates deeply with the macOS environment. It can be installed via Homebrew or built from source, requiring macOS 14.4 or later to function. Users utilize the tool to dictate text directly into other apps using voice commands and keyboard shortcuts, allowing for a hands-free writing experience that works globally across the operating system.
VoiceInk is built using Swift and leverages whisper.cpp for high-performance inference of OpenAI Whisper models. It also utilizes FluidAudio for Parakeet model implementation. The architecture focuses on local processing to ensure that audio data does not leave the device, making it suitable for users with strict privacy requirements. The application incorporates several dependencies to manage its functionality, including Sparkle for updates, KeyboardShortcuts for user customization, and SelectedTextKit for interacting with highlighted text on the screen.
By combining local AI inference with system-wide integration, VoiceInk provides a way to convert speech to text without relying on cloud-based APIs. This design ensures that the transcription process remains functional without an internet connection while maintaining data sovereignty.
This tool is positioned as a privacy-centric dictation utility specifically for the macOS ecosystem.
Join our newsletter to get shiny new open source software delivered to your inbox. Unsubscribe anytime.