MaxKB: build AI agents without code — 20K GitHub stars
MaxKB is an open-source platform for building AI agents with drag-and-drop workflows, RAG, and multi-model support. v2.7.0 just dropped with skills management and conversation sharing.
MaxKB (Max Knowledge Brain) just hit 20,000 GitHub stars and released v2.7.0 — and it lets anyone build AI-powered chatbots and agents without writing a single line of code. Think of it as a free, self-hosted alternative to paid platforms like Botpress or Voiceflow.
What MaxKB Actually Does
MaxKB is a platform where you upload your company's documents — PDFs, Word files, web pages — and it creates an AI assistant that can answer questions based on that knowledge. The technology behind this is called RAG (Retrieval-Augmented Generation), which basically means the AI looks up relevant information from your documents before answering, instead of making things up.
The key difference from ChatGPT or Claude: MaxKB runs on your own server. Your data never leaves your building. For businesses handling sensitive information — legal firms, hospitals, financial companies — that's a dealbreaker feature.
Key Numbers
- 20,400+ GitHub stars
- 6,597 commits across 58 releases
- Supports Claude, GPT, Gemini, DeepSeek, Llama, Qwen — swap models anytime
- Deploy in one Docker command
v2.7.0: Skills, Sharing, and Smarter Workflows
The March 19 release brings three major additions:
Skills Management — Your AI agents can now learn and use specific "skills" (pre-built capabilities like searching a database, calling an API, or generating a report). Instead of configuring everything manually, you assign skills and the agent figures out when to use them.
Conversation Sharing — You can now share entire chat sessions with colleagues. If your AI gives a great answer about company policy, share that conversation link instead of copy-pasting.
Knowledge Base Tagging — Documents can now be tagged and organized by category, making it easier to manage large knowledge libraries with hundreds of files.
Drag-and-Drop Workflows for Complex Tasks
MaxKB isn't just a Q&A bot. It includes a visual workflow builder where you can chain together multiple AI steps — like having one step classify a customer's question, another search relevant documents, and a third generate a personalized answer. All without code.
It also supports MCP (Model Context Protocol — the standard that lets AI tools talk to each other), meaning your MaxKB agents can connect to external tools and databases.
Who Should Care
If you're building a customer support chatbot — MaxKB lets you create one trained on your own help docs, FAQs, and product manuals. No monthly fees per conversation.
If you manage internal knowledge — HR teams, legal departments, and operations managers can create AI assistants that answer employee questions based on actual company policies, not generic ChatGPT responses.
If you're privacy-conscious — Everything runs on your own infrastructure. You can even use fully local AI models like DeepSeek or Llama, keeping all data offline.
Try It in One Command
If you have Docker installed, you can have MaxKB running in under a minute:
docker run -d --name=maxkb --restart=always -p 8080:8080 -v ~/.maxkb:/opt/maxkb 1panel/maxkb
Then open http://localhost:8080 and log in with admin / MaxKB@123.. to start building.
How It Stacks Up
Compared to commercial alternatives like Botpress (freemium, hosted) or Dify (open-source competitor with 90K+ stars), MaxKB focuses specifically on knowledge-heavy use cases — companies that need AI grounded in their own documents. While Dify is more of a general-purpose AI app builder, MaxKB's RAG pipeline and document management are more polished out of the box.
The trade-off: MaxKB's community is primarily Chinese-speaking, so English documentation is thinner. But the UI itself is fully bilingual, and Docker deployment is universal.
With 20K stars and 58 releases, MaxKB is one of the most actively maintained open-source AI agent platforms available — and it's completely free under the GPLv3 license.
Related Content — Get Started with Easy Claude Code | Free Learning Guides | More AI News
Stay updated on AI news
Simple explanations of the latest AI developments