This project provides a Telegram bot enabling users to interact with OpenAI's ChatGPT through text and voice. It's designed for individuals seeking a conversational AI interface integrated with Telegram, offering the convenience of voice input and output.

How It Works

The bot leverages LangchainJS to manage prompt construction, conversation history, and interaction with external services. It integrates with OpenAI's Whisper API for speech-to-text conversion and Play.ht for text-to-speech, facilitating seamless voice communication.

Quick Start & Requirements

• Install via npm install after cloning the repository.
• Run with npm start.
• Prerequisites: Node.js 18+, Telegram bot token, OpenAI API key (GPT-4 recommended), Play.ht API keys (User ID, Secret Key, Voice), and ffmpeg installed for voice features.
• Configuration is done via a .env file.

Highlighted Details

• Supports both text and voice interactions with ChatGPT.
• Utilizes LangchainJS for robust prompt engineering and state management.
• Integrates with Whisper API for voice input and Play.ht for voice output.
• Allows configuration of OpenAI models and user-specific access.

Maintenance & Community

The project is maintained by RafalWilinski, who also offers consulting services. Discussions can be found on Twitter and HackerNews.

Licensing & Compatibility

The repository does not explicitly state a license in the provided README. Users should verify licensing for commercial use or integration with closed-source projects.

Limitations & Caveats

Voice interaction requires ffmpeg to be installed and configured. The bot's functionality is dependent on the availability and pricing of OpenAI and Play.ht APIs.