Langflow

Langflow is a visual platform for building and deploying AI-powered agents and workflows. It provides both a drag-and-drop interface and programmatic access, turning every workflow into an API or MCP server that integrates with any application stack.

Key Features

• Visual workflow builder with drag-and-drop components
• Multi-agent orchestration with conversation management
• Support for all major LLMs and vector databases
• Interactive playground for testing flows step-by-step
• Export workflows as APIs or JSON for Python applications
• Deploy workflows as MCP servers for tool integration
• Built-in observability with LangSmith and LangFuse
• Source code access for Python customization

Installation

Install via pip:

pip install langflow -U

Or with uv:

uv pip install langflow -U

For containerized deployment:

docker run -p 7860:7860 langflowai/langflow:latest

Basic Usage

Start the Langflow server:

langflow run

Access the visual interface at http://127.0.0.1:7860. Build workflows by dragging components, configure LLM connections, and test your agents in the interactive playground.

Export your workflow as an API endpoint or integrate it programmatically:

from langflow.load import run_flow_from_json

result = run_flow_from_json(flow="path/to/flow.json", input_value="Hello, world!")

Project Details

• License: MIT
• Language: Python
• Community: 146k+ GitHub stars, active Discord server
• Deployment: Supports all major cloud platforms with official guides