How can I use multiple models in a single thread?

API
, ,
1

Hi, we’re using the Assistant API and we want to use multiple models in the same thread.

So for example, a user used to chat on model GPT-4o in ‘Thread A’, but after GPT-4o-mini was released, we wanted to use the mini model instead to save cost. How can we make the user chatting on ‘Thread A’ continue his chat on the same thread but using the mini model instead?

2

We’re also using the Assistant API on the Azure Open AI, and does anyone know how we can do the same on Azure?