Since Claude 3.5 Sonnet beat GPT-4 in June 2024, Claude has had the lead. GPT-5.5 just took it back. And Anthropic holding back Mythos is why. Look at the benchmark snapshot. GPT-5.5 beats Opus 4.7 on Terminal-Bench, BrowseComp, OSWorld, and CyberGym. But it doesn't beat Mythos. Anthropic shipped Mythos Preview on April 7 to Project Glasswing partners only. Apple, Google, Microsoft, Amazon, Nvidia, banks. Cyber defense only. The model autonomously found thousands of zero-days across every…