Skip to main content

​
LangChain

pip install langchain-openai

from langchain_openai import ChatOpenAI

llm = ChatOpenAI(
    model="orchid01",
    base_url="https://llm.orchid.ac/v1",
    api_key="orchid-your-key-here",
    temperature=0.1,
)

msg = llm.invoke("Summarise this 10-K filing...")
print(msg.content)
Do not set OPENAI_API_KEY to a real OpenAI key when using Orchid. Pass your Orchid key via api_key directly.

​
Streaming with LCEL

from langchain_core.output_parsers import StrOutputParser
from langchain_core.prompts        import ChatPromptTemplate

prompt = ChatPromptTemplate.from_messages([
    ("system", "You are a financial analyst."),
    ("user",   "{input}"),
])

chain = prompt | llm | StrOutputParser()

for chunk in chain.stream({"input": "Analyse this filing..."}):
    print(chunk, end="", flush=True)

​
LangGraph

LangGraph uses LangChain chat models as nodes. Configure ChatOpenAI as above and pass it into your graph: 
from langchain_openai          import ChatOpenAI
from langgraph.prebuilt        import create_react_agent
from langchain_core.tools      import tool

llm = ChatOpenAI(
    model="orchid01",
    base_url="https://llm.orchid.ac/v1",
    api_key="orchid-your-key-here",
)

@tool
def get_filing(ticker: str, form: str) -> str:
    """Get an SEC filing for a company"""
    # your implementation
    ...

agent = create_react_agent(llm, [get_filing])
result = agent.invoke({"messages": [("user", "Get Apple's latest 10-K")]})
No Orchid-specific graph API — standard OpenAI-style chat completions under the hood.