Migrate from gpt-3.5-turbo to gpt-4o-mini in Chat Completions API

API
1

I have been using the gpt-3.5-turbo model through the Chat Completions API.

Is migrating to the gpt-4o-mini model as simple as just changing the model reference from “gpt-3.5-turbo” to “gpt-4o-mini”?

I tried that, but when I asked “Is your model gpt-4o-mini?” the response I got was “No, my model is not GPT-4o-mini, I am based on GPT-3.5.”

What else do I need to do to get to gpt-4o-mini?

Thank you.
–Bruce

2

What is the API response JSON? It will include the model attribute. What’s the value?

You cannot expect the LLM to return sensible answers to this kind of question.

3

Unfortunately I am using a C# NuGet library, and it does not give me access to the JSON. I think you are asking for the value of “model” in the chat completion object. What the library offers is “id”, “choices”, “created”, and “usage”, but not “model”.

Is there another way I could verify the model?

4

The answer to this is yes, you are probably using the latter now, but really your logging is really poor if you cannot inspect API responses.

I would investigate ways of improving the transparency of what you are doing.

Many client libraries do provide ways to set up logging of external calls and responses.

1 Like
5

Thanks for your help. --BH

2 Likes