Flowise¶

What it is¶

Flowise is an open-source UI visual tool to build customized LLM flows. It is built on top of LangChain and allows you to create complex LLM chains and agents using a drag-and-drop interface.

What problem it solves¶

Makes it possible to build and iterate on LLM chains and agent workflows visually, without needing to write LangChain code directly.

Where it fits in the stack¶

AI & Knowledge — provides a no-code visual builder for LLM pipelines that can integrate with local Ollama models and other stack components.

Typical use cases¶

Building chatbot flows with retrieval-augmented generation
Prototyping LangChain-based agent workflows via drag-and-drop
Implementing support chatbots backed by local documentation

Strengths¶

Open-source and self-hostable
Drag-and-drop interface built on the mature LangChain ecosystem
Supports a wide range of LLM providers and vector stores

Limitations¶

Tightly coupled to LangChain, which may limit flexibility with other frameworks
Visual interface can become unwieldy for very complex flows
Debugging issues may require understanding the underlying LangChain code

When to use it¶

When you want a visual way to build and test LangChain-based LLM flows
When non-technical users need to create or modify LLM pipelines

When not to use it¶

When you need full programmatic control or want to use a framework other than LangChain
When the application is simple enough that a few lines of code suffice

Getting started¶

Installation¶

Install Flowise globally via npm and start it:

# Install Flowise globally
npm install -g flowise

# Start Flowise
npx flowise start

Minimal Example¶

Once running, you can access the UI at http://localhost:3000 to begin building your LLM flows. To interact with your flow programmatically, use the Prediction API as shown in the examples below.

curl -X POST "http://localhost:3000/api/v1/prediction/{your-chatflow-id}" \
     -H "Content-Type: application/json" \
     -d '{"question": "Hello!"}'

API examples¶

Calling a Flowise Chatflow via REST¶

You can interact with your deployed chatflows using the Prediction API, including configuration overrides.

curl -X POST "http://localhost:3000/api/v1/prediction/{your-chatflow-id}" \
     -H "Content-Type: application/json" \
     -d '{
            "question": "How do I set up a vector store in Flowise?",
            "overrideConfig": {
                "returnSourceDocuments": true
            }
         }'

Sources / references¶

Contribution Metadata¶

Last reviewed: 2026-03-03
Confidence: medium