Langflow

Overview

Langflow is a visual authoring platform for AI agents and workflows, written in Python. Every component's source is editable in place (with hot reload during local development), and a flow can be exported as a plain API or deployed directly as an MCP server, turning it into a tool any MCP client can call.

How it works

• Build flows on a drag-and-drop canvas, then test them step by step in the interactive playground.
• Orchestrate multiple agents with conversation management and retrieval built in.
• Deploy the same flow as a REST API, export it as JSON for a Python app, or expose it as an MCP server for LLM clients.
• Hook up LangSmith or LangFuse for observability.
• Langflow Desktop ships all dependencies bundled for Windows/macOS, so you don't need to manage a Python environment to get started.

Examples

• Prototyping a RAG or multi-agent flow visually, then exporting it as an MCP tool for Claude Desktop, Cursor or Windsurf.
• Running Langflow Desktop locally to experiment without any Python setup.
• Wrapping a flow as a REST API to call from an existing application.

Installation

You need Python 3.10+ and uv, or nothing at all if you use Langflow Desktop.

1. Quick start with uv:

   uv pip install langflow -U
   uv run langflow run
   

   This opens Langflow at http://127.0.0.1:7860.
2. No Python setup: download Langflow Desktop for Windows or macOS from the project's GitHub releases and run the installer instead.
3. Inside Langflow, add your LLM provider's API key under Settings.
4. Build a flow on the canvas, or start from a template, then test it in the Playground before exporting it as an API or an MCP server.