AA-PersonalityBench ranks Gemini 4 Argon most metaphorical model, GPT-3.5 Turbo least

ArtificialAnlys · x · 2026-11-17

Artificial Analysis published the "metaphorical" dimension of its AA-PersonalityBench, measuring how much models use analogies and figurative language — one of five traits in the Expressiveness index.

The gap highlights how differently models express the same ideas, with Google's lineup clustering at the top and older OpenAI models at the bottom.

Related event: Artificial Analysis Launches Personality Bench Ranking AI Models by Figurative Language(2 posts)→

Original post →

More from Models

Models channel →