A macOS speech-to-text app offering on-device AI transcription, system-wide dictation, and AI text processing with full privacy.
TypeWhisper is a macOS application that converts speech to text using on-device AI models or cloud APIs. It provides system-wide dictation, batch file transcription, and AI-powered text processing with custom prompts, all designed to keep audio data private on the user's computer.
macOS users who need reliable, private dictation and transcription, including developers, writers, and professionals who handle sensitive audio content.
Developers choose TypeWhisper for its strong privacy focus with local AI processing, extensive customization through profiles and plugins, and robust automation features like an HTTP API and CLI, all built natively for macOS.
Local speech-to-text for macOS on-device AI, fully private, optional cloud
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
Defaults to on-device AI models like WhisperKit and Parakeet, ensuring voice data stays local unless cloud APIs are explicitly used, as emphasized in the README's philosophy.
Offers eight transcription engines including local options (e.g., WhisperKit for 99+ languages) and cloud APIs (Groq, OpenAI), providing flexibility for speed and accuracy needs.
Enables per-app and per-website overrides for language, engine, prompts, and hotkeys, with URL pattern matching and priority rules for tailored dictation workflows.
Includes a local HTTP API and CLI tool for scripting transcription, along with a plugin SDK to extend functionality with custom engines or LLM providers.
Requires macOS 14.0+, with key features like Apple Intelligence and SpeechAnalyzer needing macOS 26+, locking out users on older or incompatible systems.
On-device processing consumes significant RAM and storage, with the README recommending 8-16GB+ for larger models, making it unsuitable for low-spec Macs.
Demands granting microphone and accessibility permissions, downloading AI models, and configuring profiles or plugins, which can be barrier for casual users.