langchain-ai/langgraph

Health v2 · maintenance only

Adoption not part of the score

40503 stars · 6830 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

LangGraph is a low-level orchestration framework and runtime for building, managing, and deploying long-running, stateful AI agents in Python. It lets developers mix deterministic hand-coded steps with LLM-driven agentic steps in a graph, with durable execution, human-in-the-loop interrupts, and short- and long-term memory.

Use cases

• build a stateful multi-step LLM agent
• orchestrate multi-agent workflows with shared state
• add human approval checkpoints to agent runs
• build agents that survive failures and resume execution
• create a customer support chatbot with long-term memory
• mix deterministic business logic with LLM tool-calling steps
• build a RAG agent with controllable control flow

When to choose

• you need fine-grained control over agent control flow and state
• you are building long-running, production-grade agents that need durability and memory
• you want human-in-the-loop oversight of agent execution
• you need multi-agent orchestration with shared, inspectable state

When to avoid

• you just want a simple prebuilt agent loop with minimal setup - use LangChain agents or Deep Agents instead
• your application is a simple LLM call chain without state or branching
• you need a JavaScript/TypeScript-first codebase - use LangGraph.js instead

Facets

framework · maturity stable

agent-framework workflow-automation llm-inference rag chatbot state-management large-language-models artificial-intelligence developer-tools python cross-platform llm-agents graph-orchestration durable-execution human-in-the-loop multi-agent langchain-ecosystem stateful-agents ai-agents retrieval-augmented-generation

Member repositories

For agents

markdown · JSON · MCP: product_card(name="langchain-ai/langgraph")