# Getting full API effectiveness without upgrading Python SDK

**URL:** <https://community.openai.com/t/getting-full-api-effectiveness-without-upgrading-python-sdk/507855>\
**Category:** API\
**Created:** [November 16, 2023, 4:47pm UTC](https://community.openai.com/t/getting-full-api-effectiveness-without-upgrading-python-sdk/507855 "2023-11-16T16:47:47Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![domenicthomas](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/domenicthomas/32/217424_2.png) [@domenicthomas](https://community.openai.com/u/domenicthomas)\
**Post date:** [November 16, 2023, 4:47pm UTC](https://community.openai.com/t/getting-full-api-effectiveness-without-upgrading-python-sdk/507855/1 "2023-11-16T16:47:47Z")

</div>

So I have been unable to make the new SDK compatible with my existing Python/Langchain Chatbot. Whenever I load it I get errors, and frankly I’m too much of a noob to figure it out without a lot of pain trying to resolve the dependencies.

I’m referencing the new model (gpt-4-1106-preview) in my code where it’s appropriate, but I’m not sure I’m getting its full capabilities due to being still on the openai 0.28.1 SDK.

I have been testing it by asking it questions, like when is the last update of your training data, and when I use Openai’s ChatGPT chatbot I get April 2023, but when I run the same question through my chatbot, I still get Sept 2021 or whatever it was.

Is it just that you have to have the new SDK implemented? Or do I not get the same answer in my chatbot because OpenAI runs some kind of secret sauce in their ChatGPT chatbot?
