MaxKB - Build Knowledge Base with AI

Can MaxKB crawl and ingest web pages automatically into the RAG pipeline?

Yes. MaxKB supports automatic web crawling alongside document uploads to populate your knowledge base. The platform then handles text splitting and vectorization on the ingested content, which feeds into the retrieval-augmented generation pipeline to ground agent responses in your source material.

Which language models can MaxKB run, and do I need to use OpenAI or Claude?

MaxKB is model-agnostic and works with DeepSeek, Llama, Qwen, OpenAI, Claude, and Gemini. You can deploy private language models on your own infrastructure or connect to commercial APIs, so you are not locked into any single vendor.

Does MaxKB let me connect to third-party systems and databases without writing code?

Yes. MaxKB advertises zero-coding integration for connecting existing systems to the platform. The agentic workflow engine and function library support MCP tool-use, which lets you wire up external APIs and databases through the interface rather than custom code.

What happens if I want to fork MaxKB or run it on my own servers?

MaxKB is open-source with over 20,600 stars on GitHub, so you can fork the codebase and deploy it to your own infrastructure. This gives you full control over your data and workflow logic without relying on a hosted service.