not all models support tool calling

you need to read the docs of the provider and AI sdk's to see which model supports tool calling and how

but right now if I'm not mistaken regardless of the model we can't do tool calling with generateObject

I think you can, not sure

let me read the docs

yea sadly its only in generate and stream text

you could however, do 1 call to generateText with tool calling

and pass the output to generateObject to get desired output

yeah that what I found I just want to undertand the thinking behind the AI sdk api design.

Unfortunately, I don't think that is also possible because you want the LLM to choose which tool to call, and if you cannot pass tools, it means the LLM will not get the tools' definition in the API code.

I meant to first call generateText with tools and then pass the output of that call to generateText to get the output you need

I am following this approach, just without the tools part

That would only work if you already know which tool you want the LLM to call. But, say you have five tools and you want the LLM to decide which tool to call, get the result, and decide what is the next tool to call. For the LLM to do that, it needs to get the list of tools that it has available.