This project provides a web interface for interacting with ChatGPT, built using Express and Vue3. It targets developers and users who want a self-hosted, customizable ChatGPT experience, offering flexibility in API usage and deployment.

How It Works

The application supports two primary methods for accessing ChatGPT: the official OpenAI API (using gpt-3.5-turbo) and an unofficial proxy API that leverages web access tokens. The unofficial method bypasses Cloudflare and offers a "smarter" response, though it's noted as less reliable and dependent on third-party servers. Configuration is managed via environment variables, allowing users to specify API keys, access tokens, models, and proxy settings.

Quick Start & Requirements

• Install: pnpm install (backend), pnpm bootstrap (frontend)
• Run: pnpm start (backend), pnpm dev (frontend)
• Prerequisites: Node.js (^16, ^18, ^19), PNPM. Docker is also supported.
• Configuration: Requires OPENAI_API_KEY or OPENAI_ACCESS_TOKEN set in service/.env.
• Docs: Introduction

Highlighted Details

• Supports dual models (OpenAI API and unofficial proxy).
• Offers extensive configuration options via environment variables for API endpoints, models, proxies, and security.
• Provides Docker images and docker-compose configurations for easy deployment.
• Includes Nginx configuration snippets to prevent web crawlers.

Maintenance & Community

The project is actively maintained by Chanzhaoyu. Acknowledgements are given to JetBrains for providing an open-source license. Links to contributing guides are available.

Licensing & Compatibility

• License: MIT
• Compatibility: The MIT license permits commercial use and linking with closed-source projects.

Limitations & Caveats

The unofficial proxy method is described as "relatively unreliable" and dependent on third-party servers. Users are warned about potential risks when using access tokens with third-party proxies. The project also mentions a commit message verification process for contributors.