Description

A comprehensive head-to-head benchmark where Anthropic's Claude Opus 5, Opus 4.8, and Fable 5 were tasked with one-shotting four distinct games: a physics sandbox, Tetris, Breakout, and a raindrop simulation. The video evaluates each model's coding logic and the 'feel' of the final artifacts through blind scoring, while also tracking the relative token costs of each one-shot generation using Claude Code.

Descriptions, tags, and model credits may be AI-generated or inferred from public sources and can be incomplete or wrong. Learn more.