NeoSyntropyProxy

Client / LangGraph

Use NeoSyntropy inside LangGraph

Drop a NeoSyntropy-backed chat model into a StateGraph while LangGraph keeps ownership of state, edges, and node execution.

Install the adapter

Terminalbash
pip install "neosyntropy[langgraph]"

Create the model client

Pythonpython
import os
from neosyntropy.integrations.langgraph import NeoSyntropyChatModel

llm = NeoSyntropyChatModel(
    api_key=os.environ["NEOSYNTROPY_API_KEY"],
    model="default",
    project_id=os.environ.get("NEOSYNTROPY_PROJECT_ID"),
)

reply = llm.invoke("Summarize the customer request.")
print(reply.content)

Wire it into a StateGraph

Pythonpython
from typing import TypedDict
from langgraph.graph import END, StateGraph

class State(TypedDict):
    request: str
    summary: str

def summarize(state: State):
    answer = llm.invoke(state["request"])
    return {"summary": str(answer.content)}

graph = StateGraph(State)
graph.add_node("summarize", summarize)
graph.set_entry_point("summarize")
graph.add_edge("summarize", END)
app = graph.compile()

result = app.invoke({"request": "Customer asks to move invoice 2048."})