Satire: If Claude's Literary Model Names Keep Going, Brace for 'Cinematic Universe'

A short humor piece riffs on Anthropic’s literary naming convention for Claude models — Haiku, Sonnet, Opus, and now Fable — by extrapolating the scheme to its absurd conclusion. The author imagines a full ‘literary stack’ of future models, each name paired with a deadpan jab at LLM behavior: ‘Marginalia’ offers unprompted code commentary, ‘Abstract’ summarizes reasoning it never did, and ‘Treatise’ leaves citations as an exercise for the reader.

The sharpest gags target cost and verbosity. ‘Fable (xhigh)’ is described as a ‘bankruptcy speedrun,’ ‘Saga’ adds extra meandering, and ‘Cinematic Universe (Director’s Cut)’ delivers the same answer with 42% more tokens. The piece lands as gentle mockery of two industry-wide trends: increasingly baroque model naming and the way larger, pricier models often trade efficiency for sprawl.

It’s a one-minute read with no technical payload, but it captures a real frustration among developers — that model tiers have become harder to parse, and that ‘more capable’ frequently just means ‘more tokens billed.’