Claude API LangChain Example

Use Claude with LangChain in Python. Install langchain-anthropic, build chains, use LCEL, add memory, and stream responses — working 2026 examples.

LangChain's langchain-anthropic package wraps the Anthropic SDK so Claude fits natively into any LangChain pipeline — chains, agents, retrievers, and memory.

Installation

pip install langchain langchain-anthropic

Basic chat call

from langchain_anthropic import ChatAnthropic
from langchain_core.messages import HumanMessage

llm = ChatAnthropic(model="claude-sonnet-4-6")  # reads ANTHROPIC_API_KEY from env

response = llm.invoke([HumanMessage(content="Explain LangChain in two sentences.")])
print(response.content)

LCEL chain (prompt → model → parser)

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

prompt = ChatPromptTemplate.from_messages([
    ("system", "You are a concise technical writer."),
    ("human", "{question}")
])

llm = ChatAnthropic(model="claude-sonnet-4-6", max_tokens=512)

chain = prompt | llm | StrOutputParser()

answer = chain.invoke({"question": "What is prompt caching?"})
print(answer)

Streaming with LCEL

for chunk in chain.stream({"question": "Explain agentic AI step by step."}):
    print(chunk, end="", flush=True)

Conversation with memory

from langchain.memory import ConversationBufferMemory
from langchain.chains import ConversationChain

llm = ChatAnthropic(model="claude-sonnet-4-6", max_tokens=1024)
memory = ConversationBufferMemory()

conversation = ConversationChain(llm=llm, memory=memory)

print(conversation.predict(input="Hi, my name is Alex."))
print(conversation.predict(input="What's my name?"))  # Claude remembers "Alex"

RAG chain (retrieval-augmented generation)

from langchain_anthropic import ChatAnthropic
from langchain_core.prompts import ChatPromptTemplate
from langchain_core.output_parsers import StrOutputParser
from langchain_core.runnables import RunnablePassthrough

# Assume 'retriever' is any LangChain retriever (Chroma, FAISS, Pinecone, etc.)
prompt = ChatPromptTemplate.from_template("""
Answer the question based only on the following context:
{context}

Question: {question}
""")

llm = ChatAnthropic(model="claude-sonnet-4-6")

rag_chain = (
    {"context": retriever, "question": RunnablePassthrough()}
    | prompt
    | llm
    | StrOutputParser()
)

answer = rag_chain.invoke("What are Claude's context window limits?")

LangChain vs raw Anthropic SDK

Use the raw SDK when you want minimal dependencies and direct control. Use LangChain when you need retrievers, multi-provider portability, or LangSmith tracing. Estimate costs for either approach at Claude API Cost Calculator.

Frequently asked questions

Free tools

More examples