LangGraph is a low-level orchestration framework and runtime environment designed for building reliable, stateful AI agents.
1.1 API Configuration
LangGraph builds on LangChain and connects to LLMs through ChatOpenAI. To use an OpenAI-compatible service, set openai_api_base to your endpoint.
Basic Configuration
from langchain_openai import ChatOpenAI
llm = ChatOpenAI(
openai_api_base="https://api.rygen.io/v1",
api_key="sk-your-rygen-api-key",
model="your-model-name"
)Environment Variables
import os
from langchain_openai import ChatOpenAI
os.environ["OPENAI_API_BASE"] = "https://api.rygen.io/v1"
os.environ["OPENAI_API_KEY"] = "sk-your-rygen-api-key"
llm = ChatOpenAI(model="your-model-name")1.2 Using in LangGraph
from langgraph.prebuilt import create_react_agent
from langchain_openai import ChatOpenAI
# Configure LLM
llm = ChatOpenAI(
openai_api_base="https://api.rygen.io/v1",
api_key="sk-your-rygen-api-key",
model="your-model-name",
temperature=0.4
)
# Create agent
agent = create_react_agent(llm, tools=[])
# Run agent
result = agent.invoke({"messages": [("user", "Hi there!")]})
print(result)