For the first time in a long time, it feels like OpenAI has a better model than Anthropic. GPT-5.6 has everything you'd want: - It works really hard - The $20 plan has tons of usage - It's at the frontier of all benchmarks The last time I felt this way was GPT-4, which led all models from March to June 2023, until the launch of Claude 2. Since then, I've been Claude pilled. But GPT-5.6-sol is giving me the "big model" feel. Rumors say Fable sits on a 1T parameter pre-train while GPT-5.6-sol…