Topic Group: Common Migration Errors
langchain_openai and langchain_community, and ensure you call the OpenAI or Anthropic clients with the current SDK patterns and model names like gpt-4o or claude-3-5-sonnet-20241022.code_error langchain_openai and langchain_community and switch to the new client instantiation and method calls as per SDK v1+.Why this happens
LangChain has updated its SDK and import paths, deprecating older modules like langchain.llms and langchain.chat_models. Using these deprecated imports or old client patterns causes import errors or silent failures. Additionally, calling OpenAI or Anthropic APIs with outdated methods like openai.ChatCompletion.create() or missing the system= parameter for Claude leads to runtime errors or unexpected behavior.
Example broken code:
from langchain.llms import OpenAI
client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])
response = client.chat.completions.create(
model="gpt-3.5-turbo",
messages=[{"role": "user", "content": "Hello"}]
)
print(response.choices[0].message.content)ModuleNotFoundError: No module named 'langchain.llms'
The fix
Use the updated imports from langchain_openai and langchain_community. Instantiate the client directly from the provider SDK (e.g., OpenAI or anthropic.Anthropic) with api_key from os.environ. Use current model names like gpt-4o or claude-3-5-sonnet-20241022. For Claude, use the system= parameter instead of a system message in the messages array.
This works because it aligns with the latest SDK v1+ patterns and avoids deprecated modules and methods.
from openai import OpenAI
import os
client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])
response = client.chat.completions.create(
model="gpt-4o",
messages=[{"role": "user", "content": "Hello"}]
)
print(response.choices[0].message.content)
import anthropic
client_anthropic = anthropic.Anthropic(api_key=os.environ["ANTHROPIC_API_KEY"])
message = client_anthropic.messages.create(
model="claude-3-5-sonnet-20241022",
max_tokens=1024,
system="You are a helpful assistant.",
messages=[{"role": "user", "content": "Hello"}]
)
print(message.content[0].text)Hello Hello
Preventing it in production
Implement automated tests to verify your LangChain imports and client calls after upgrades. Use dependency management tools to lock versions and monitor changelogs for breaking changes. Add retry logic for transient API errors and validate model names and parameters before runtime. Consider using CI pipelines to catch deprecated usage early.
Key Takeaways
- Always update LangChain imports to langchain_openai and langchain_community to avoid import errors.
- Use the latest OpenAI and Anthropic SDK client patterns with correct model names like gpt-4o and claude-3-5-sonnet-20241022.
- Validate and test your AI integration code after upgrades to catch deprecated usage early.