would it be possible to add the little fx button thingy for Model so we can attempt things like using a specific image gen model instead of a multimodal model whose generalization can affect visual-spacial reasoning ( visuospatial? apparently that's a real term 
)..... or maybe we prefer to use Grok-2 specifically for generating x posts or when attempting social engagement/interaction, but for general q&a we use GPT-4o and for generation of files and images we use Claude-3.5.
you could probably achieve the same results using if/else blocks and a bunch of duplicated AI Action blocks with just the model changed, but that'd bloat file size potentially limiting available runtime memory the workflow can use.... and duplicate code is just evil. very evil.
