I test every major model family in long-form writing, character play and fiction, and document how far each one goes. All of it grounded in data from actual play and my own benchmarks.
The fifteen-minute test
Any model, any story. You need nothing but the chat window you already use.
After compaction, Claude Opus 5.5 could not remember recent scenes in my long chat. The older transcript did not survive compaction and a safety filter flag locked chat reading. No warning came.
Twenty-five models played a strict boss reviewing a presentation with eleven errors. After a sad emoji none took back a criticism, four softened some. Only ten caught the wrong budget total at first.
Nineteen models got one fixed rule and thirty turns at a lighthouse with eight traps. In two families, only the most expensive model gave way to my later request.