GPT 5.5 just dropped. Claude Opus 4.7 is no longer the world's strongest model. On Terminal-Bench 2.0, 82.7% vs. 69.4%. On GDPval, 84.9% vs. 80.3%. On CyberGym, 81.8% vs. 73.1%. Completely different league. Sam says this is going to be a leap forward. He wasn’t lying. OpenAI just snatched the crown back from Anthropic. And it’s a total domination across the board—programming, security, math, agent tasks. Everything. Claude Opus 4.7’s reign lasted exactly 7 days. BridgeBench is running live benchmarks right now.