What it is
Flowise is an open-source platform for building AI agents, RAG chatbots and multi-agent flows on a visual node canvas. It's built on LangChain, supports dozens of models and vector databases, and every flow is available as an API or an embeddable chat widget for your site. Install it with npm or Docker, or use Flowise Cloud.
How to use it
- Build a chatbot on your documents by wiring loaders, embeddings and a vector database.
- Design agents with tools, memory and human approval steps in Agentflow.
- Publish each flow as an API or an embeddable chat widget.
- Self-host it with npm or Docker.
- Try models from OpenAI, Anthropic, Google or Ollama by swapping a single node.
What it's good for
- A university adding an assistant to its website that answers questions about policies and procedures.
- A dev team prototyping a RAG agent before moving it into code.
- A company with sensitive data running local models through Ollama on its own server.
Pros
- Open source under Apache 2.0, except the enterprise module
- Every flow is available as an API and a chat widget
- Supports local models like Ollama so data stays in-house
- Free cloud plan to try it without installing anything
Cons
- Complex flows call for a working grasp of LLMs, embeddings and RAG
- Interface and docs are English-only
- The free cloud plan is very limited: 2 flows and 100 predictions a month
