[Open Source] A Complete Open-Source AI Assistant Solution Based on AI Large Language Model APIs
![[Open Source] A Complete Open-Source AI Assistant Solution Based on AI Large Language Model APIs](https://yy123.ink/api/img/1782602267057-df0nuz.jpg)
1. Project Overview
ChatGPT-Plus
This is a complete system that includes both a frontend chat application and a backend administration system. The system features user authentication. You can use it yourself or deploy it to provide ChatGPT services directly to end users.
ChatGPT-PLUS is a complete open-source AI assistant solution built on APIs for large language models. It comes with an operations and management dashboard and is ready to use out of the box. It integrates large language models from multiple platforms, including OpenAI, Azure, ChatGLM, iFlytek Spark, and ERNIE Bot. It also integrates AI image generation features powered by MidJourney and Stable Diffusion. Key features include:
-
A complete open-source system, with both the frontend application and backend administration system ready to use out of the box.
-
Built on WebSocket, delivering a smooth typewriter-style experience.
-
Includes various pre-trained role-based assistants, such as Xiaohongshu Content Writer, English Translation Expert, Socrates, Confucius, Steve Jobs, Weekly Report Assistant, and more. Easily meet various chatting and application needs.
-
Supports multiple large language models, including OpenAI, Azure, ERNIE Bot, iFlytek Spark, and Tsinghua ChatGLM.
-
Supports integrated MidJourney / Stable Diffusion AI image generation, ready to use out of the box.
-
Supports using a personal WeChat QR code as a payment channel for recharges, without requiring a corporate payment gateway.
-
Integrates Alipay payment functionality and supports multiple membership plans and prepaid card purchases.
-
Integrates a plugin API that can be used with large language model function capabilities to develop powerful plugins. Built-in plugins include Weibo trending topics, Toutiao news, daily morning news, and AI image generation functions.
2. Open Source License
Licensed under MIT
Usage Instructions
-
This project is licensed under the MIT License. All source code is freely available and can be used for personal learning or commercial purposes.
-
For commercial use, copyright information must be retained. Please comply with this requirement voluntarily. Ensure all usage is legal and compliant. You are solely responsible for any consequences arising from operation of the project. The author assumes no liability.
3. Screenshots
Feature Screenshots
PC Chat Interface

AI Conversation Interface

MidJourney Professional Drawing Interface

Stable Diffusion Professional Drawing Page


Gallery of Artwork

AI Application List

Membership Recharge

Automatic Function Calling Plugins


Admin Console




Mobile Web Page




4. Feature Overview
The following features have been implemented in the new version:
-
Introduced a user system with new user registration and login functionality.
-
Redesigned the chat page to provide a chatting experience consistent with the official ChatGPT version.
-
You can select a chat role and model when creating a conversation.
-
Added chat settings, allowing users to import their own API KEY.
-
Chat history is saved, with support for conversational context.
-
Refactored the backend management module into a more user-friendly and extensible management system.
-
Integrated the ip2region component to record users' login IP addresses and locations.
-
Supports conversation search and filtering.
-
Supports account top-ups via WeChat Pay.
TODOLIST
-
Integrated the Midjourney AI drawing API.
-
Developed a mobile chat page.
-
Integrated WeChat payment collection.
-
Supports ChatGPT function calling, enabling plugins through functions.
-
Developed a desktop application.
-
Developed a mobile App client.
-
Alipay payment functionality.
-
Supports AI-powered Q&A based on a knowledge base.
-
Membership referral program.
-
WeChat Pay functionality.
5. Tech Stack
Project Technical Architecture
The new version has undergone a major refactoring of both the frontend and backend. The backend still uses the Gin Web framework, but the author has integrated the fx dependency injection framework, resulting in a very clean backend application structure that is particularly suitable for secondary development. In addition, MySQL has replaced leveldb for data storage. Since the system targets C-end users, many business data queries and statistics will be required in the future, and leveldb is no longer sufficient.
Gin + fx + MySQL
After version 3.0, API implementations in other languages, such as PHP and Java, will be added gradually. Considering the author's limited time and resources, the api directory has already been added. Interested contributors are welcome to claim and implement the language they are most familiar with.
The frontend framework remains:
Vue3 + Element-Plus
The page styles of both the frontend and backend have been completely redesigned, with almost all page styling code rewritten. The existing logic code has been retained, since the functionality has not changed significantly.
This refactoring and redesign was primarily intended to prepare for future feature expansion.
See the README.md documentation for details