🧠 MaxKB: Free Open-Source AI Agent Platform for Enterprise Knowledge Bases

MaxKB is a free, open-source AI knowledge base platform featuring RAG retrieval, multi-model support, visual workflows, and one-click Docker deployment for enterprises.

Open-source, efficient, and model-agnostic — empower every enterprise with its own "super brain"!
In today's rapidly evolving AI landscape, how can enterprises efficiently leverage large language models to build proprietary knowledge bases? MaxKB (Max Knowledge Brain) is the answer! As an open-source knowledge base Q&A system built on LLMs and RAG technology, it tackles the pain points of high technical barriers, costly deployment, and lengthy iteration cycles that enterprises face when adopting AI. Whether for intelligent customer service, internal knowledge management, or academic research, MaxKB delivers an out-of-the-box Q&A experience to help enterprises embrace the AI era.

✨ Core Feature Highlights

📚 RAG (Retrieval-Augmented Generation)

• Support direct document upload (TXT, PDF, DOCX, Markdown, etc.) or automatic web crawling, with automatic text splitting and vectorization.
• Reduce LLM "hallucinations" through retrieval-augmented generation, improving answer accuracy and reliability.
• Optimized knowledge base segmentation with manual chunking rule adjustment to ensure critical information is never lost.

🔧 Flexible Workflow Orchestration

• Built-in visual workflow engine supporting complex AI process design (multi-turn dialogue, data queries, tool invocation).
• New "Follow-up Question" feature: AI predicts user follow-up questions and auto-generates quick reply options to boost interaction efficiency.

🌐 Model-Agnostic & Broad Compatibility

• Supports local private models (DeepSeek-R1, Llama 3, Qwen 2), major domestic models (Tongyi Qianwen, Baidu Qianfan, Kimi, etc.), and international models (OpenAI, Gemini, etc.).
• Switch models without code changes — avoid vendor lock-in.

🔗 Seamless Third-Party Integration

• API interfaces and embeddable components allow zero-code integration into WeCom, DingTalk, WeChat Official Accounts, and more.
• Support conversation history and custom UI styling for enhanced user experience.

🆚 Comparison with Similar Tools — Why Choose MaxKB?

🎯 Advantage Summary

• Low Barrier: Non-technical staff can quickly build knowledge bases through the UI.
• Flexibility: Workflows support custom AI behavior for complex scenarios (customer service, data analysis).
• Security: Local deployment ensures data privacy, with LDAP/OIDC authentication support.

🛠️ Pro Tips: Unlock Advanced Features

🚀 Optimize Q&A Performance

• Adjust text chunk size and overlap in knowledge base settings to balance retrieval precision and speed.
• Use the "Hit Test" feature to verify knowledge base coverage and supplement missing documents.

💡 Workflow Automation Examples

• Intelligent Customer Service: Configure "Question Understanding → Knowledge Retrieval → Answer Generation → Follow-up Recommendation" to reduce manual intervention.
• Data Querying: Connect database tool nodes to dynamically query inventory or business data.

🔌 API Integration Example

import requests
base_url = "http://localhost:8080/api"
auth = ("admin", "your_password")

# Send question to application
response = requests.post(
    f"{base_url}/apps/your_app_id/chat",
    json={"question": "Which models does MaxKB support?", "stream": False},
    auth=auth
)
print(response.json()["answer"])

Note: The API enables batch Q&A processing or integration into business systems.

📥 Installation & Deployment Guide

⚙️ Environment Requirements

• OS: Ubuntu 22.04 / CentOS 7 / Windows 10+.
• Specs: 2-core CPU, 4GB RAM, 100GB disk recommended.
• Dependencies: Docker (recommended) or offline installer.

🐳 Deployment Steps

• One-Click Docker Deployment

# Linux
docker run -d --name=maxkb --restart=always -p 8080:8080 \
  -v ~/.maxkb:/var/lib/postgresql/data \
  registry.fit2cloud.com/maxkb/maxkb

# Windows
docker run -d --name=maxkb --restart=always -p 8080:8080 \
  -v C:/maxkb:/var/lib/postgresql/data \
  registry.fit2cloud.com/maxkb/maxkb

Data volume mounts ensure persistent storage

• 🔐 Initial Configuration

Visit http://your-server-ip:8080 and log in with username admin / password MaxKB@123...

• Change the password on first login, then add models and knowledge bases.
• 🤖 Model Integration (Ollama Local Model Example)

Install Ollama and download a model (e.g., Llama 2):

ollama run llama2

• Fill in the Ollama API address in MaxKB's "Model Settings" (e.g., http://127.0.0.1:11434)
• 💽 Offline Deployment

Use offline installer packages in intranet environments, or deploy quickly via the 1Panel App Store.

💎 Conclusion

With its model neutrality, out-of-the-box usability, and flexible extensibility, MaxKB is the ideal choice for enterprises building intelligent knowledge bases. Whether you're a technical team or business staff, MaxKB lowers the barrier to AI adoption and enables intelligent knowledge management. Just one Docker command is all it takes to start building your enterprise's "super brain"!

🚀 Get Started Now

• Explore the source on GitHub
• Check the official documentation for detailed setup examples
• Watch the Bilibili video tutorial to get started quickly

Note: This is the English translation of the original Chinese version.