Type: Agent framework.
Low-level orchestration framework for custom long-running, stateful agents.
Capabilities: Durable execution, human-in-the-loop control, persistence, streaming and agent orchestration.
MIT-licensed core framework. Model APIs, hosting and optional LangSmith products have separate terms and costs; core library licensing does not imply free managed deployment. Requires application code and tool/model configuration.
Official sources checked on 23 October 2026: https://github.com/langchain-ai/langgraph https://docs.langchain.com/oss/python/langgraph/overview https://github.com/langchain-ai/langgraph/blob/main/LICENSE
Getting started
Install the Python library with pip install -U langgraph. Define state, graph nodes and edges, configure model/tool calls and persistence, then compile and invoke your graph.