LangGraph: Step by Step (9 Modules)
State, routing, tool-calling agents, persistence, human approval, streaming, subgraphs, and deployment.
9 lessons · about 78 minutes of reading
Log in to track your progress through this course.
Start lesson 1- 1
LangGraph Module 1: Introduction and Your First Graph
Learn what LangGraph is, how state, nodes, and edges fit together, and build and run your first graph.
8 min read
- 2
LangGraph Module 2: State Schemas and Reducers
Control how node updates merge into state with reducers, use message state, and separate input and output schemas.
8 min read
- 3
LangGraph Module 3: Conditional Edges, Routing, and Loops
Branch and loop your graph with conditional edges and Command, so the path depends on the state.
8 min read
- 4
LangGraph Module 4: Building an LLM Agent with Tools from Scratch
Build the classic tool-calling agent loop as a graph: an LLM node, a tool node, and a conditional edge.
10 min read
- 5
LangGraph Module 5: Persistence, Checkpointers, and Threads
Save graph state between runs with checkpointers, resume conversations by thread ID, and inspect history.
9 min read
- 6
LangGraph Module 6: Human-in-the-Loop with Interrupts
Pause a graph to ask a person for approval or input, then resume exactly where it stopped.
9 min read
- 7
LangGraph Module 7: Streaming Output and Debugging
Stream state updates, tokens, and custom progress events from a graph, and learn to trace and debug runs.
8 min read
- 8
LangGraph Module 8: Parallel Branches and Subgraphs
Run steps in parallel, merge their results with reducers, and compose larger systems from smaller graphs.
9 min read
- 9
LangGraph Module 9: Local Server, Studio, and Deployment
Run your graph with the LangGraph dev server, debug visually in Studio, call it over an API, and plan for production.
9 min read