Guides
Cookbooks
JS Integration Langchain
This is a Jupyter notebook

Cookbook: Langchain Integration (JS/TS)

This is a cookbook with examples of the Langfuse Integration for Langchain (JS/TS).

Follow the integration guide (opens in a new tab) to add this integration to your Langchain project. The integration also supports Langchain Python.

Setup

Initialize the Langfuse client with your API keys from the project settings in the Langfuse UI and add them to your environment.

import { CallbackHandler } from "npm:langfuse-langchain"
const langfuseLangchainHandler = new CallbackHandler({
    publicKey: "",
    secretKey: "",
    baseUrl: "https://cloud.langfuse.com",
    flushAt: 1 // cookbook-only: do not batch events, send them immediately
})

Langchain interfaces

Langfuse supports the following Langchain JS interfaces

  • invoke
  • stream

For this section we will use a very simple example prompt (from Langchain JS docs (opens in a new tab)) and ChatOpenAI. Langfuse works with any model.

import { ChatOpenAI } from "npm:@langchain/openai"
import { PromptTemplate } from "npm:@langchain/core/prompts"
 
const model = new ChatOpenAI({});
const promptTemplate = PromptTemplate.fromTemplate(
  "Tell me a joke about {topic}"
);

invoke

import { RunnableSequence } from "npm:@langchain/core/runnables";
 
const chain = RunnableSequence.from([promptTemplate, model]);
 
const res = await chain.invoke(
    { topic: "bears" },
    { callbacks: [langfuseLangchainHandler] }
);
 
console.log(res.content)
Why did the bear wear a fur coat to the BBQ?
Because it was grizzly cold outside!

stream

const chain = promptTemplate.pipe(model);
const stream = await chain.stream(
    { topic: "bears" },
    { callbacks: [langfuseLangchainHandler] }
);
for await (const chunk of stream) {
  console.log(chunk?.content);
}
Why
 did
 the
 bear
 bring
 a
 flashlight
 to
 the
 party
?
 


Because
 he
 wanted
 to
 be
 the
 "
bear
"
 of
 the
 light
!

Explore the trace in Langfuse

In the Langfuse interface, you can see a detailed trace of all steps in the Langchain application.

Langfuse Trace

Was this page useful?

Questions? We're here to help

Subscribe to updates