An internal system to try out different models for different tasks, and I wonder if there's any efficient way to query an API endpoint or a feature included in the AI SDK to figure out which AI model is able to work with what kind of tools. For example, whether it supports web search, or if it can handle PDF files or images. Also, information about whether it supports reasoning or not would be very helpful. My goal would be that I don't have to do anything on the developer's side to have new models supported through Vercel AI Gateway or Open Router automatically available in my application, capable of doing at least the basic stuff if they support it.