OpenWhispr/openwhispr
Privacy-first, cross-platform voice-to-text dictation with AI agents and meeting transcription, offering local and cloud processing options.
Core Features
Quick Start
npm install && npm run devDetailed Introduction
OpenWhispr is a privacy-first, cross-platform desktop application designed for voice-to-text dictation, note-taking, and AI-powered interactions. It offers a unique blend of local, on-device processing using models like Whisper and NVIDIA Parakeet, ensuring audio never leaves the user's device, alongside optional cloud integration for speed. Key features include global hotkey dictation, AI agent capabilities, and advanced meeting transcription with live speaker diarization and Google Calendar integration. Positioned as an open-source alternative to commercial dictation software like WisprFlow and Granola, OpenWhispr emphasizes user privacy with no data collection or telemetry, making it an ideal tool for professionals handling sensitive information across macOS, Windows, and Linux.