Skip to main content

Your First Chain

A minimal chain: a prompt template feeds a chat model, whose output feeds a parser. The | operator wires them together.

from langchain.chat_models import init_chat_model
from langchain_core.output_parsers import StrOutputParser
from langchain_core.prompts import ChatPromptTemplate

prompt = ChatPromptTemplate.from_template(
    "Explain {topic} in one sentence for a beginner."
)
model = init_chat_model("gpt-4o-mini", model_provider="openai")
parser = StrOutputParser()

chain = prompt | model | parser

result = chain.invoke({"topic": "vector embeddings"})
print(result)

What's happening​

  1. prompt takes a dict ({"topic": "..."}) and fills the template, producing a PromptValue.
  2. model takes that PromptValue, calls the LLM, and returns an AIMessage.
  3. parser takes the AIMessage and extracts just the string content.

| doesn't call anything — it builds a RunnableSequence object. Nothing runs until .invoke(...) is called on the composed chain.

Every piece in this chain implements the same Runnable interface — that's why | works uniformly regardless of what's on either side of it. That interface, and the rest of what a Runnable can do beyond invoke, is covered next.

See also​

  • installation.md, keys-and-config.md — getting to the point this snippet assumes.
  • Core Primitives → Runnables & LCEL (next section) — the full Runnable contract this chain relies on.