Build a Chatbot With Flowise in 10 Minutes
Flowise is a visual builder for LLM applications, chatbots, agents, and retrieval pipelines, using a drag-and-drop node editor instead of writing orchestration code by hand.
1. Install Flowise
- From your dashboard, go to Marketplace.
- Find Flowise under AI Agents and click Install.
- Open it at its subdomain once it's running.
2. Start a new chatflow
- Click Chatflows → Add New.
- You'll land on a blank canvas. This is where you assemble your chatbot from nodes.
3. Add the core nodes
A basic conversational chatbot needs three pieces:
- Chat Model — search the node panel for your preferred provider (OpenAI, Anthropic, or a local model) and drag it onto the canvas. Add your API key as a credential the same way n8n does.
- Memory — add a Buffer Memory node so the bot remembers earlier messages in the conversation.
- Conversation Chain — connect the Chat Model and Memory nodes into a Conversation Chain node. This is what actually runs the conversation loop.
4. Connect the nodes
Drag a line from the Chat Model's output to the Conversation Chain's model input, and from Memory's output to the Conversation Chain's memory input.
5. Test it
Click the chat icon in the top right to open the built-in chat panel. Send a message, if everything's connected correctly, you'll get a response within a few seconds.
6. Go further
From here, Flowise supports adding document retrieval (RAG) by connecting a vector store node, or turning the flow into an agent with access to tools. Both are drag-and-drop additions to the same canvas.
7. Deploy it
Once you're happy with the flow, use the API tab to get an embeddable widget or a REST endpoint you can call from your own app.