How to beginner · 3 min read

How to migrate from langchain to langchain-openai

Q: How to migrate from langchain to langchain-openai

To migrate from langchain to langchain-openai, update your imports to use langchain_openai modules and instantiate models like ChatOpenAI directly from langchain_openai. Replace deprecated imports and patterns with the new SDK v1+ client usage for clean, production-ready code.

Quick answer

To migrate from langchain to langchain-openai, update your imports to use langchain_openai modules and instantiate models like ChatOpenAI directly from langchain_openai. Replace deprecated imports and patterns with the new SDK v1+ client usage for clean, production-ready code.

PREREQUISITES

Python 3.8+
OpenAI API key (free tier works)
pip install langchain-openai openai>=1.0

Setup

Install the langchain-openai package and ensure your OPENAI_API_KEY is set in your environment variables.

Run:

bash

pip install langchain-openai openai>=1.0

Step by step

Replace old langchain imports with langchain_openai. Instantiate ChatOpenAI with your API key from os.environ. Use the generate method with a prompt to get completions.

python

import os
from langchain_openai import ChatOpenAI

# Initialize the ChatOpenAI client
client = ChatOpenAI(
    model_name="gpt-4o",
    openai_api_key=os.environ["OPENAI_API_KEY"]
)

# Define a prompt
prompt = "Write a Python function to reverse a string."

# Generate a completion
response = client.generate(prompt)

print(response.generations[0].text)

output

def reverse_string(s):
    return s[::-1]

Common variations

Use different models by changing model_name, e.g., gpt-4o-mini.
For embeddings, import and use OpenAIEmbeddings from langchain_openai.
Async usage is supported by calling async methods on ChatOpenAI.

python

from langchain_openai import OpenAIEmbeddings

embeddings = OpenAIEmbeddings(
    model_name="o1-mini",
    openai_api_key=os.environ["OPENAI_API_KEY"]
)

vector = embeddings.embed_query("Hello world")
print(vector)

output

[0.123, -0.456, ...]  # vector floats

Troubleshooting

If you get import errors, verify you installed langchain-openai and not the deprecated langchain package.
Ensure your OPENAI_API_KEY is set correctly in your environment.
Check that you are using the correct model names like gpt-4o as older names are deprecated.

✅

Key Takeaways

Use langchain_openai imports instead of deprecated langchain modules.
Instantiate ChatOpenAI with your API key from os.environ for secure usage.
Update model names to current ones like gpt-4o to avoid compatibility issues.

Verified 2026-04 · gpt-4o, gpt-4o-mini, o1-mini

Verify ↗