Control over scope of context provided to Assistant?

I don’t see any way (yet) in the Assistants API to control the scope — i.e. how much stuff — to include as context. The docs simply say that the API “intelligently chooses which context from the thread to include when calling the model”. I gather that there is no user control over this. but it feels like an essential feature, since it has a big impact on cost.

1 Like