VoiceInk is a native macOS application that transcribes spoken words to text using local AI models with near-instant processing.
The tool solves the problem of voice-to-text transcription on macOS without requiring subscriptions or cloud services. It uses local AI models running entirely on the user's device, ensuring all audio processing stays offline and private. The application detects which app or website you are using and automatically applies pre-configured transcription settings for that context, adapting its behavior based on screen content. It includes a personal dictionary feature to train the AI on custom terminology and industry-specific language, configurable keyboard and mouse shortcuts for recording, and multiple AI-powered modes optimized for different writing styles.
VoiceInk suits developers and users who prioritize privacy and want to avoid subscription-based voice transcription services. It works best for those already invested in the macOS ecosystem and willing to purchase a license for automatic updates and priority support. The README positions it as an alternative to commercial services like Superwhisper and Wispr Flow, emphasizing its offline-first approach and lack of subscription requirements as key differentiators.
The project was open-sourced after five months of dedicated development but is not currently accepting pull requests; contributors can fork the repository for personal modifications or report bugs and suggest features through issues. Documentation for building from source and contributing guidelines are provided. The maintainer offers a free trial through the project website and accepts purchases to fund full-time development and ongoing feature improvements.
