How to beginner · 3 min read

How to use AgentOps with LangChain

Q: How to use AgentOps with LangChain

Use agentops.init() to enable automatic tracing of LangChain calls by patching the OpenAI client. Then run your LangChain chains or agents normally; all LLM interactions will be tracked by AgentOps without extra code changes.

Quick answer

Use agentops.init() to enable automatic tracing of LangChain calls by patching the OpenAI client. Then run your LangChain chains or agents normally; all LLM interactions will be tracked by AgentOps without extra code changes.

PREREQUISITES

Python 3.8+
OpenAI API key
pip install langchain openai agentops

Setup

Install the required packages and set your environment variables for OPENAI_API_KEY and AGENTOPS_API_KEY. AgentOps automatically instruments OpenAI SDK calls used by LangChain.

bash

pip install langchain openai agentops

output

Collecting langchain
Collecting openai
Collecting agentops
Successfully installed langchain openai agentops

Step by step

This example shows how to initialize AgentOps to auto-track LangChain calls using the OpenAI client. Run a simple LangChain ChatOpenAI chain and see AgentOps capture the session.

python

import os
import agentops
from langchain_openai import ChatOpenAI

# Initialize AgentOps with your API key to enable automatic tracing
agentops.init(api_key=os.environ["AGENTOPS_API_KEY"])

# Create a LangChain ChatOpenAI client
chat = ChatOpenAI(model="gpt-4o-mini", temperature=0)

# Run a simple prompt
response = chat.invoke([{"role": "user", "content": "What is RAG in AI?"}])

print("Response:", response.content)

output

Response: RAG stands for Retrieval-Augmented Generation, a technique that combines retrieval of relevant documents with generative models to improve accuracy and context.

Common variations

Use agentops.init() without parameters to auto-patch OpenAI if AGENTOPS_API_KEY is set in env vars.
For async LangChain usage, AgentOps tracks calls the same way.
Works with other OpenAI-based LangChain models like ChatOpenAI or OpenAIEmbeddings.

python

import asyncio
import agentops
from langchain_openai import ChatOpenAI

agentops.init()  # Uses AGENTOPS_API_KEY from env

async def main():
    chat = ChatOpenAI(model="gpt-4o-mini", temperature=0)
    response = await chat.ainvoke([{"role": "user", "content": "Explain LangChain."}])
    print("Async response:", response.content)

asyncio.run(main())

output

Async response: LangChain is a Python framework for building applications with large language models, enabling chaining, memory, and integrations.

Troubleshooting

If AgentOps does not track calls, ensure AGENTOPS_API_KEY is set correctly in your environment.
Verify you are using the OpenAI SDK v1+ and LangChain v0.2+ for compatibility.
Restart your Python environment after installing agentops to ensure patching works.

✅

Key Takeaways

Initialize AgentOps with agentops.init() before running LangChain to enable automatic tracing.
AgentOps patches the OpenAI SDK used by LangChain, requiring no code changes to your chains or agents.
Ensure environment variables AGENTOPS_API_KEY and OPENAI_API_KEY are set for authentication.
AgentOps supports both synchronous and asynchronous LangChain usage seamlessly.
Use the latest LangChain and OpenAI SDK versions for best compatibility with AgentOps.

Verified 2026-04 · gpt-4o-mini

Verify ↗