LLMRouterChain Class — langchain Architecture

Architecture documentation for the LLMRouterChain class in llm_router.py from the langchain codebase.

Class python

Entity Profile

Dependency Diagram

graph TD
  6583c2eb_e73e_04a5_2dfe_d3e26b9ec1eb["LLMRouterChain"]
  8c933306_16e9_7f72_a392_80e52d7a7371["RouterChain"]
  6583c2eb_e73e_04a5_2dfe_d3e26b9ec1eb -->|extends| 8c933306_16e9_7f72_a392_80e52d7a7371
  12b08e15_dda7_34a6_0911_10fbdd2f2a7b["llm_router.py"]
  6583c2eb_e73e_04a5_2dfe_d3e26b9ec1eb -->|defined in| 12b08e15_dda7_34a6_0911_10fbdd2f2a7b
  56745521_2936_9cc1_c805_d317353974fe["_validate_prompt()"]
  6583c2eb_e73e_04a5_2dfe_d3e26b9ec1eb -->|method| 56745521_2936_9cc1_c805_d317353974fe
  eba221c9_aa38_e298_cf9a_bcbc4aed2278["input_keys()"]
  6583c2eb_e73e_04a5_2dfe_d3e26b9ec1eb -->|method| eba221c9_aa38_e298_cf9a_bcbc4aed2278
  39ce415b_b0c9_e015_5e2b_1408f4edea5e["_validate_outputs()"]
  6583c2eb_e73e_04a5_2dfe_d3e26b9ec1eb -->|method| 39ce415b_b0c9_e015_5e2b_1408f4edea5e
  87ccce0a_1b7e_e8ff_7889_acaf7978975c["_call()"]
  6583c2eb_e73e_04a5_2dfe_d3e26b9ec1eb -->|method| 87ccce0a_1b7e_e8ff_7889_acaf7978975c
  f3be5ec5_70d6_5fa7_102d_c7ef16f74255["_acall()"]
  6583c2eb_e73e_04a5_2dfe_d3e26b9ec1eb -->|method| f3be5ec5_70d6_5fa7_102d_c7ef16f74255
  7e98a1df_1041_92f3_e31b_2dfd178eb89d["from_llm()"]
  6583c2eb_e73e_04a5_2dfe_d3e26b9ec1eb -->|method| 7e98a1df_1041_92f3_e31b_2dfd178eb89d

Relationship Graph

Source Code

libs/langchain/langchain_classic/chains/router/llm_router.py lines 33–164

class LLMRouterChain(RouterChain):
    """A router chain that uses an LLM chain to perform routing.

    This class is deprecated. See below for a replacement, which offers several
    benefits, including streaming and batch support.

    Below is an example implementation:

        ```python
        from operator import itemgetter
        from typing import Literal
        from typing_extensions import TypedDict

        from langchain_core.output_parsers import StrOutputParser
        from langchain_core.prompts import ChatPromptTemplate
        from langchain_core.runnables import RunnableLambda, RunnablePassthrough
        from langchain_openai import ChatOpenAI

        model = ChatOpenAI(model="gpt-4o-mini")

        prompt_1 = ChatPromptTemplate.from_messages(
            [
                ("system", "You are an expert on animals."),
                ("human", "{query}"),
            ]
        )
        prompt_2 = ChatPromptTemplate.from_messages(
            [
                ("system", "You are an expert on vegetables."),
                ("human", "{query}"),
            ]
        )

        chain_1 = prompt_1 | model | StrOutputParser()
        chain_2 = prompt_2 | model | StrOutputParser()

        route_system = "Route the user's query to either the animal "
        "or vegetable expert."
        route_prompt = ChatPromptTemplate.from_messages(
            [
                ("system", route_system),
                ("human", "{query}"),
            ]
        )


        class RouteQuery(TypedDict):
            \"\"\"Route query to destination.\"\"\"
            destination: Literal["animal", "vegetable"]


        route_chain = (
            route_prompt
            | model.with_structured_output(RouteQuery)
            | itemgetter("destination")
        )

        chain = {
            "destination": route_chain,  # "animal" or "vegetable"
            "query": lambda x: x["query"],  # pass through input query
        } | RunnableLambda(
            # if animal, chain_1. otherwise, chain_2.
            lambda x: chain_1 if x["destination"] == "animal" else chain_2,
        )

        chain.invoke({"query": "what color are carrots"})

        ```
    """

    llm_chain: LLMChain
    """LLM chain used to perform routing"""

    @model_validator(mode="after")
    def _validate_prompt(self) -> Self:
        prompt = self.llm_chain.prompt
        if prompt.output_parser is None:
            msg = (
                "LLMRouterChain requires base llm_chain prompt to have an output"
                " parser that converts LLM text output to a dictionary with keys"
                " 'destination' and 'next_inputs'. Received a prompt with no output"