Is there a way to switch the API key to the new GPT 4o model? How does it work? I saw the api was available today and wasn’t sure if the api key was ready for it or not.
GPT-4o is one of many AI models available on the API for use.
The model name is one of several parameters when making an API call to get a response.
The API has been around over two years. You pay for the actual usage by input and output of data.
Documentation link on the side of the forum has a quckstart and overview, but it doesn’t go super-beginner.
If you’re using 3.5-turbo and NOT instruct, you should be okay just replacing model name with the latest.
Do you know how to swap from 3.5 turbo instruct to 4o. As you suggest changing the model name doesn’t work.
Welcome to the forum.
That’s because the Instruct and Chat models are different.
https://platform.openai.com/docs/models/model-endpoint-compatibility
You’ll need to check the docs on migrating from Completion (Legacy) to Chat Completion… it’s a bit of work, but it’s better for some things and worth it…
Let us know if you get stuck.
Thanks I’ll head over there.
1 Like