epicenter

• Purpose: Whispering is an open-source transcription application that provides global speech-to-text functionality, powered by OpenAI's Whisper model for high accuracy. It's designed to make dictation seamless with features like keyboard shortcuts and automatic copy/paste.

• Key Features:

  • Global Transcription: Access speech-to-text anywhere via keyboard shortcuts or quick button clicks.
  • Cross-Platform Support: Available as a desktop app (Windows, macOS, Linux) and Chrome extension.
  • Chat Integration: Adds recording buttons to ChatGPT and Claude for direct voice input.
  • Local Data Storage: Prioritizes privacy by storing recordings and transcriptions locally (IndexedDB).
  • Configurable Transcription Services: Supports external services (OpenAI, Groq) or local options like faster-whisper-server.

• Target Audience: Users needing accurate, on-the-fly transcription across applications, with a focus on privacy-conscious individuals and those integrating voice input into workflows (e.g., chat interfaces).

• Framework Support: Built with Svelte 5 for the web/desktop UI (via SvelteKit and Tauri) and React for the Chrome extension (using Plasmo).

• Data Privacy: Emphasizes local storage; only recordings are sent to external services if explicitly chosen by the user.

For demos and more details, visit the web app or GitHub repository .