WhisperChain is a Python application that streamlines workflows by converting speech to text, then using AI to clean up filler words and refine the output. It targets users seeking to accelerate tasks through voice input, offering features like real-time transcription, automatic clipboard integration, and global hotkey support.

How It Works

The system leverages Whisper.cpp for efficient, real-time speech recognition. Transcribed audio is then processed by LangChain for AI-powered cleanup, removing filler words and improving clarity. A FastAPI server handles audio streaming and processing via WebSockets, while a Streamlit UI provides an interactive interface for prompts and history. Global hotkeys trigger recording, and cleaned text is automatically copied to the clipboard.

Quick Start & Requirements

• Install system dependencies (macOS): brew install ffmpeg portaudio
• Install project: pip install whisperchain
• Requires Python 3.8+ and an OpenAI API Key.
• Configuration can be set via environment variables, a .env file, or ~/.whisperchain/.env.
• Start application: whisperchain
• Streamlit UI: streamlit run src/whisperchain/ui/streamlit_app.py
• Official Docs: [Not explicitly linked, but project structure implies local documentation]

Highlighted Details

• Real-time speech recognition via Whisper.cpp.
• AI-powered transcription cleanup using LangChain.
• Global hotkey support for voice control.
• Automatic clipboard integration of cleaned transcriptions.

Maintenance & Community

• Project appears to be actively developed by a single primary author (chrischoy).
• No explicit community links (Discord, Slack) or roadmap are provided in the README.

Licensing & Compatibility

• The LICENSE file is mentioned but not provided in the README content.
• Compatibility for commercial use or closed-source linking is not specified.

Limitations & Caveats

The project requires an OpenAI API key, incurring potential costs. System dependencies are specific to macOS, and Windows/Linux support is not detailed. The project is primarily command-line driven with an optional Streamlit UI.