I definitely agree. Even on the Professional plan there's just an overwhelming amount of LLMs to choose from. Sure I can try them one-by-one but that's tedious. Alternatively, instead of being sorted by the AI provider, maybe they could be sorted by use-case and then by quality (in that category)/cost. Such as built-in categories for general writing, technical writing, creative writing, scientific research, general research, ideation, coding, etc.
As I write this it almost sounds like an argument for Kagi expanding the built-in Custom Assistants. Either way, a more practical system for finding the models that work best for my specific use-cases would be a huge improvement. I understand the difficulties here: this is a really subjective thing, among other things. But it would be a huge value-add for me.
