Michael R. Bock @michaelrbock.com · Feb 20

4/ Full updated rankings (using strict scoring where every line must be correct): Opus 4.6: 52.94% GPT-5 w/ Search: 41.67% GPT-5.2 Pro: 41.18% Sonnet 4.6: 37.25% <-- new Gemini 3 Pro: 36.27% Opus 4.5: 36.27% GPT-5.2: 33.82% Gemini 2.5 Pro: 32.35% 7 months ago, 32% was SOTA.

1 likes 0 replies

?