This project provides an unofficial, out-of-the-box implementation of ChatGPT for QQ and WeChat, enabling users to integrate advanced AI capabilities into their daily messaging. It targets individuals who may have difficulty accessing ChatGPT directly, offering text and image generation, and even voice replies for QQ.

How It Works

The bot leverages the mirai (for QQ) and itchat4j-uos (for WeChat) libraries to log into and monitor messages on these platforms. Incoming messages are then forwarded to the OpenAI API for processing via the openai-java library. Responses from OpenAI, including text and image generation (DALL-E 3) or text-to-speech, are then relayed back to the user within the chat interface. The architecture supports multiple API keys for load balancing and allows for session resets and custom system prompts to define bot personality.

Quick Start & Requirements

• Install: Clone the repository and configure application.yml.
• Prerequisites: An OpenAI account with an API key, a QQ or WeChat account. A proxy is recommended for users in regions blocking OpenAI API access.
• Configuration: Edit application.yml with API keys, proxy settings, desired models (GPT-3.5/GPT-4, DALL-E 3, TTS-1/TTS-1-HD), and QQ/WeChat account details.
• Run: Execute the application.
• Docs: OpenAI API Keys, OpenAI Models

Highlighted Details

• Supports text dialogue with GPT-3.5/GPT-4.
• Integrates DALL-E 3 for AI image generation (returns URL for WeChat).
• Offers AI voice replies using TTS models for QQ.
• Allows configuration of multiple OpenAI API keys for load balancing.
• Supports session reset and custom system prompts for personality.

Maintenance & Community

The project is maintained by a university student and welcomes community contributions and ideas. The latest update was on January 15, 2024. Contact is available via email (ashinnotfound@qq.com).

Licensing & Compatibility

The project's licensing is not explicitly stated in the README. However, it relies on other open-source projects which may have their own licenses. Commercial use compatibility is not specified.

Limitations & Caveats

AI image and voice features are limited on WeChat (image returns URL only, voice not supported). The project notes occasional unresponsiveness for image/voice features, suggesting re-login or cache clearing. QQ login stability has been an issue, with QR code scanning offered as a workaround. The project is unofficial and may be affected by changes in QQ, WeChat, or OpenAI's API policies.