AI Model Comparison

A benchmark comparison showing AI language models competing in Tetris gameplay. Gemini 3 Flash leads with a 66% win rate, followed by Gemini 3 Pro at 64%. GPT-5.2 achieves 54%, Claude Opus 4.5 reaches 52%, Claude Sonnet 4 scores 41%, while Grok 4.1 Fast Reasoning trails at 19%. Results are based on head-to-head matches between Show more