ChatWiki: Open-Source AI Knowledge Base Q&A System

ChatWiki is an open-source AI Q&A platform designed for building intelligent knowledge bases. It leverages Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), and GraphRAG knowledge graphs to handle complex data processing and model interactions. Companies, universities, and government agencies can utilize ChatWiki to rapidly deploy private AI Q&A systems. The platform supports diverse document formats—including DOCX, Excel, PDF, and OFD—and is accessible directly through web browsers. It is compatible with over 20 leading models, such as DeepSeek, Qwen, and Doubao.

Users are encouraged to start with ChatWiki’s cloud version to test ideas and interaction patterns before committing to a standalone deployment. This approach significantly reduces initial trial costs. Visit chatwiki.com to explore the official demo and experience the interface via the ChatWiki WebApp or dedicated client.

By importing existing organizational data, you can build a specialized knowledge base that serves as the foundation for the AI bot’s responses. This allows you to establish a dedicated AI Q&A system in minutes. The platform supports a wide array of global models, including DeepSeek R1, Doubao Pro, Qwen Max, OpenAI, and Claude.

ChatWiki features a flexible workflow configuration engine that enables multi-step task automation. You can customize Q&A logic and data routing to align with specific business requirements, facilitating intelligent collaboration in complex environments. Additionally, the chatbot can connect directly to your internal business systems.

Deployment is highly versatile; ChatWiki can be embedded into websites, desktop clients, web apps, WeChat mini-programs, WeChat public accounts, WeChat customer service, Douyin enterprise accounts, Kuaishou accounts, video channels, and via API calls. It is designed to cover every multi-device scenario required by modern enterprises.

The system automatically extracts embedded images from uploaded PDFs, Word documents, and other knowledge base files. When a user’s question matches content containing an image, the bot provides a combined text-and-image response, ensuring a more comprehensive answer.

ChatWiki also includes integrated enterprise help center tools. You can publish your knowledge base as a public-facing documentation site, complete with SEO optimization and traffic analytics. This helps organizations build professional, branded customer support portals with ease.

To improve accuracy, ChatWiki analyzes user queries in real time and automatically clarifies vague or incomplete questions. Through intent recognition and semantic association, it converts raw input into precise search commands, significantly boosting retrieval precision and the relevance of the answers provided.

Based on semantic analysis, the system generates a "Questions You Might Ask" list to guide users. Administrators can manually maintain a list of high-frequency FAQs, and the recommendation logic adapts based on individual user history to ensure higher hit rates and a better overall experience.

For security, ChatWiki offers enterprise-grade multi-level access control. Administrators can assign roles such as admin, editor, or read-only, making it ideal for managing sensitive data within team collaborations. With support for over 20 global models, setup is as simple as configuring your model’s API key.

Data can be imported through several methods: automatic segmentation, QA-based segmentation, manual entry, or CSV upload. The system preprocesses, vectorizes, or splits text into QA pairs automatically. It supports smart segmentation for Word, Excel, PPT, PDF, OFD, Markdown, and other formats.

The visual interface is designed for simplicity. With just a few clicks, you can configure an AI Q&A bot and its associated knowledge base without needing deep technical expertise.

ChatWiki allows you to store all data in a local database protected by multi-layered security protocols, including encrypted data transfers, strict access controls, and detailed audit logs. This ensures that sensitive information remains safe and compliant with data privacy regulations.

Local Installation

To install ChatWiki locally, prepare a Linux server with internet access. The minimum recommended specifications are 4 CPU cores and 16 GB of RAM.

ChatWiki Community Edition runs on Docker. If Docker is not yet installed on your system, you can install it using the following command:

sudo curl -sSL https://get.docker.com/ | CHANNEL=stable sh

Once Docker is configured, follow these steps:

• Clone the ChatWiki project repository. From GitHub: git clone https://github.com/zhimaAi/chatwiki.git From Gitee: git clone [email protected]:zhimaAi/chatwiki.git

• Navigate to the chatwiki/docker directory and use Docker Compose to build and start the project: docker compose up -d

• Access the system via your server's IP and port (ensure port 18080, or your configured ${CHAT_SERVICE_PORT}, is open). Default account: admin Default password: chatwiki.com@123

If you encounter issues during installation or deployment, the support team is available for assistance. Detailed documentation is available covering one-click deployments, offline installations, non-Docker setups, Baota deployment (contributed by the community), model provider configuration, local model integration, and API key management.

Technical Architecture

The system utilizes JWT authentication and Casbin for robust permission management. The backend architecture includes:

• Database: PostgreSQL 16 with pgvector and zhparser.
• Cache: Redis 5.0.
• Web Server: Nginx.
• Asynchronous Queue: NSQ.
• Process Management: Supervisor.
• Frontend: Vue.js.
• Backend: Go and Python.

ChatWiki supports a broad range of models, including OpenAI, Google Gemini, Claude 3, Tongyi Qianwen, Wenxin Yiyan, iFlytek Spark, Baichuan, and Tencent Hunyuan.