What is MaxKB?

MaxKB, short for Max Knowledge Brain, is an open-source platform for building enterprise-grade AI agents, knowledge-base assistants, and Retrieval-Augmented Generation (RAG) applications. It helps organizations connect large language models with private documents, business data, workflows, functions, and external tools.

The platform is designed for use cases such as AI customer service, internal knowledge bases, employee assistants, document question answering, academic research, and educational applications. MaxKB can be self-hosted with Docker, allowing organizations to retain control over their data and deployment environment.

Key Features

• RAG knowledge bases: Upload local files or collect documents from online sources. MaxKB automatically splits, processes, and vectorizes the content for knowledge-based question answering.
• AI agent workflows: Build multi-step AI processes with a visual workflow engine, reusable functions, conditional logic, and tool integrations.
• MCP support: Connect agents with compatible Model Context Protocol tools and external services.
• Multiple AI models: Use DeepSeek, Llama, Qwen, OpenAI, Claude, Gemini, MiniMax, and other supported private or public models.
• Multimodal capabilities: Create applications that work with text, images, audio, and video.
• No-code integration: Add MaxKB assistants to websites and existing business systems without developing a complete AI interface from scratch.
• Self-hosted deployment: Run MaxKB on your own infrastructure with Docker for greater control over privacy, configuration, and data storage.

What Can You Build with MaxKB?

MaxKB can be used to create AI assistants that answer questions using company documents, product manuals, policies, research papers, training materials, or support documentation. Its workflow and tool-use capabilities also allow agents to perform structured tasks instead of only generating conversational responses.

Common MaxKB use cases include:

• Enterprise knowledge-base assistants
• AI customer-support agents
• Internal employee help desks
• Document and PDF question answering
• Research and education assistants
• Website chatbots
• Workflow-based AI automation
• Agents connected to MCP tools

Supported AI Models

MaxKB is model-agnostic and can connect to both locally deployed and cloud-hosted models. Supported options include DeepSeek, Llama, Qwen, OpenAI models, Anthropic Claude, Google Gemini, MiniMax, and other compatible model providers.

This flexibility allows teams to select models based on performance, cost, privacy, language support, and infrastructure requirements.

Install MaxKB with Docker

Run the following command to start a MaxKB container:

docker run -d --name=maxkb --restart=always -p 8080:8080 -v ~/.maxkb:/opt/maxkb 1panel/maxkb

After the container starts, open the following address in your browser:

http://your_server_ip:8080

Default administrator credentials:

• Username: admin
• Password: MaxKB@123..

Change the default password immediately after signing in, especially when deploying MaxKB on a public server.

Users in China who cannot pull the Docker image can follow the MaxKB offline installation guide.

MaxKB Screenshots

Technical Stack

MaxKB uses the following primary technologies:

• Frontend: Vue.js
• Backend: Python and Django
• LLM framework: LangChain
• Database: PostgreSQL and pgvector
• Deployment: Docker

License

MaxKB is open-source software distributed under the GNU General Public License version 3 (GPLv3). Review the GPLv3 license before using or distributing modified versions of the software.