Looks like everyone is catching up to "opus" level intelligence. GLM 5.2 continues to remain extremely competitive as Grok 4.5 and muse spark spend billions for somewhat identical performance.