Token & Trinket is a game where one model plays the shopkeeper (in this case Sonnet 3.7), managing store finances, purchasing, inventory, pricing, debt payments etc.

Another model (Opus) plays gamemaster, choosing from a set of events that affects the economics and building a narrative around it.

gpt-image api paints a picture each day.

This is not an eval, but if it were it'd be one I'd approve of because the models seem to have a lot of fun.