GLM 5.1 vs Gemini-3.8-flash tested on visualization algorithms in LLM Arena

NickPassig · x · 2026-09-26

NickPassig tested GLM 5.1 against Gemini-3.8-flash in LLM AI Arena on visualization algorithms (flocking, patterns, etc.), concluding "both are smart and I'm dumb." Details are thin, but the model names are notable—Gemini-3.8-flash may be an unconfirmed test codename.

Original post →

More from Models

Models channel →