LangGraph’s powerful state management— including short- & long-term memory, human-in-the-loop, and time travel — just got even more flexible.

With the new Functional API, you can now use these capabilities without explicitly defining a graph. Whether you're using LangGraph within a different framework or as a standalone tool, this API makes it easier than ever to build AI workflows.

Why use the LangGraph Functional API:

• Works anywhere – Use LangGraph’s features in any app, no graph syntax required

• Seamless state management – Keep track of user inputs, workflow progress, and historical context

• Human-in-the-loop support – Pause workflows, collect feedback, and resume dynamically

• Streaming-first – Stream updates in real-time for responsive AI interactions

• Graph + Functional API = Best of both worlds – Mix and match paradigms to build robust AI systems

Learn more in our blog post: https://blog.langchain.dev/introducing-the-langgraph-functional-api/

Watch our video on the Functional API for Python: