AI Agent Orchestration Framework
30.4k 2026-04-26
langchain-ai/langgraph
A low-level orchestration framework for building, managing, and deploying resilient, stateful AI agents using a graph-based approach.
Core Features
Durable Execution: Agents persist through failures and automatically resume from their last state.
Human-in-the-Loop: Allows for human inspection and modification of agent state during execution.
Comprehensive Memory: Supports both short-term working memory and long-term persistent memory for agents.
LangSmith Integration: Provides deep visibility, tracing, and debugging tools for complex agent behaviors.
Production Deployment: Offers scalable infrastructure for deploying stateful, long-running agent systems.
Quick Start
pip install -U langgraphDetailed Introduction
LangGraph is a foundational, low-level orchestration framework designed for developing and deploying resilient, long-running, and stateful AI agents. By modeling agent logic as a graph, it enables robust workflows that can persist through failures, incorporate human oversight, and maintain comprehensive memory across sessions. It integrates seamlessly with the LangChain ecosystem, offering tools for enhanced debugging via LangSmith and production-ready deployment, making it ideal for complex AI applications requiring advanced state management and reliability.