LangGraph Module 1: Introduction and Your First Graph

Step 1: Learn the three core ideas

State is the data shared by the whole graph. Nodes are Python functions that read the state and return updates to it. Edges decide which node runs next. You define these, compile the graph, and then invoke it with an initial state.

Step 2: Install LangGraph

LangGraph installs on its own and works without LangChain, though the documentation uses LangChain components for models and tools in later modules.

pip install -U langgraph

Step 3: Define the state and nodes

State is described with a TypedDict. Each node takes the state and returns a dictionary containing only the keys it wants to change; LangGraph merges that into the state. Nodes should return updates, not modify the state in place.

from typing_extensions import TypedDict

class State(TypedDict):
    text: str

def shout(state: State) -> dict:
    return {"text": state["text"].upper()}

def add_exclamation(state: State) -> dict:
    return {"text": state["text"] + "!"}

Step 4: Connect them and compile

Add the nodes to a StateGraph, connect them with edges starting at START and ending at END, then call compile. You must compile a graph before you can use it.

from langgraph.graph import StateGraph, START, END

builder = StateGraph(State)
builder.add_node("shout", shout)
builder.add_node("add_exclamation", add_exclamation)

builder.add_edge(START, "shout")
builder.add_edge("shout", "add_exclamation")
builder.add_edge("add_exclamation", END)

graph = builder.compile()

Step 5: Run it and see how it relates to LangChain

invoke takes the initial state and returns the final state. LangChain's create_agent is built on LangGraph and returns a compiled graph, so everything you learn here applies to those agents too, and an agent can even be used as a node inside a larger graph.

print(graph.invoke({"text": "hello"}))
# {'text': 'HELLO!'}