Overview:
OpenWhispr is an open-source, cross-platform desktop application for voice-to-text dictation, meeting transcription, and note-taking. It positions itself as a free alternative to WisprFlow and Granola. The application allows users to press a hotkey, speak, and have their words appear at the cursor. It offers a choice between fully private offline transcription using local speech-to-text engines (such as Whisper and NVIDIA Parakeet) where audio does not leave the device, or cloud processing for potentially faster results. It is available for macOS, Windows, and Linux.
Core Features:
Voice dictation: Dictate into any application using a global hotkey with automatic text pasting at the cursor.
AI agent: Interact with models like GPT-5, Claude, Gemini, Groq, or local models through a named voice assistant.
Meeting transcription: Automatically detect calls on platforms like Zoom, Microsoft Teams, and FaceTime, featuring live speaker diarization, voice fingerprinting, and Google Calendar integration.
Local speaker diarization: Perform on-device speaker labeling with voice fingerprint recognition across meetings, requiring no cloud service.
Notes: Create, organize, and search notes with features including folders, semantic search, cloud sync, and AI actions.
Public API & MCP: Manage notes and transcriptions programmatically or connect an AI assistant via the Model Context Protocol (MCP) server.
Use Cases:
Distraction-free writing: Dictate text into any field or application on the desktop without switching windows, using a global hotkey.
Automated meeting capture: Transcribe calls from common video conferencing platforms with automatic speaker identification and labeling.
Personal note management: Create, organize, and retrieve notes using folders and semantic search, with optional cloud sync.
Voice-controlled AI interaction: Use a hands-free voice assistant to interface with various AI models, including local options for offline use.
Why It Matters:
As an open-source alternative to proprietary tools like WisprFlow and Granola, OpenWhispr provides full transparency and user control over data processing. Its architecture allows users to keep audio data entirely on-device by using local transcription and AI models, a capability not always available in comparable commercial software. The application is built on a modular tech stack with a public API and an MCP server, enabling developers to extend its functionality or integrate it into custom workflows.
