GLM 5.2 Preset Miscalculates by 2.667x

LiorOnAI · x · 2026-07-12

A hands-on test tasked four frontier models with building a KV-cache debugger, demanding exact formulas without cutting corners.

The results revealed:

The original poster highlighted the irony that the cheapest model was the sole one to make a mathematical error.

Original post →

More from Models

Models channel →