# Flowise

> Build LLM apps and AI agents visually with a drag-and-drop canvas

- Category: App UIs & Builders / Low-Code & No-Code Builders
- Language: TypeScript
- Repository: https://github.com/FlowiseAI/Flowise
- Website: https://flowiseai.com
- Canonical page: https://ai-tldr.dev/tools/flowise/

## In short

Flowise is an open-source tool for building LLM apps and AI agents without writing much code. You work on a visual canvas, dragging in nodes for models, prompts, memory, tools, and data sources, then wiring them together to form a flow. It sits in the Low-Code & No-Code Builders category of the App UIs & Builders tools and is open-source, written in TypeScript.

## Overview

Flowise is an open-source tool for building LLM apps and AI agents without writing much code. You work on a visual canvas, dragging in nodes for models, prompts, memory, tools, and data sources, then wiring them together to form a flow.

It runs as a Node.js app with three parts in one repository: a server for the API logic, a React user interface, and a library of third-party node integrations. Once a flow is ready, you can use it through an API endpoint, embed it as a chat widget, or self-host it on your own infrastructure.

## What it does

- Visual drag-and-drop canvas for designing LLM chains and multi-step agent flows
- Large library of built-in nodes for models, prompts, memory, tools, and data loaders
- Turns each finished flow into an API endpoint and an embeddable chat widget
- Self-host anywhere, with ready-made deployment guides for AWS, Azure, GCP, and more
- Runs locally in minutes via npm or as a Docker container
- Apache 2.0 licensed open-source codebase you can extend with custom nodes

## Getting started

Flowise needs Node.js 20 or newer. The fastest way to try it is the npm package, which starts a local server you open in your browser.

1. Install Flowise globally

   Install the Flowise CLI package using npm.

```bash
npm install -g flowise
```

2. Start the app

   Launch Flowise, then open http://localhost:3000 in your browser to reach the visual builder.

```bash
npx flowise start
```

3. Or run with Docker

   Prefer containers? Build the image locally and run it on port 3000.

```bash
docker build --no-cache -t flowise .
docker run -d --name flowise -p 3000:3000 flowise
```

4. Build from source for development

   To work on Flowise itself, clone the repo and use PNPM to install, build, and start the monorepo.

```bash
git clone https://github.com/FlowiseAI/Flowise.git
cd Flowise
pnpm install
pnpm build
pnpm start
```

## When to use it

- Prototype a chatbot or RAG assistant over your own documents without writing backend code
- Build multi-step AI agents that call tools and APIs, then expose them as an endpoint for your app
- Let non-developers on a team assemble and test LLM workflows on a shared visual canvas
- Self-host an internal AI app on your own cloud to keep data and keys inside your infrastructure