@LuminaXspace It just too slow and expensive (plus inefficient) to make its intelligence worth it. Great at medium effort, but cant justify much above that
@bridgemindai Looking into the CursorBench data, Sonnet 5 at medium effort is actually very impressive. Smashes Sonnet 4.6 Max and does it cheaper. But Extra High and Max make no sense for the new model.
@aiedge_ Then why it feel so much more advanced? It feels like you could completely refresh your software in a few hours with Fable 5, but with GPT 5.5 it feels like you have to go component by component.