Anthropic's Fable 5.2 Snatches the Crown

🔥 What happened Anthropic shipped Claude Fable 5.2 and Mythos 5.2, reclaiming the #1 spot on the Intelligence Index at 65 — just ten days after Gemini 4 Carbon briefly took it. Per-token price dropped 20%, but cost per task jumped to $26.50 because the model spins up subagents and writes thousands of test cases instead of twenty. 💡 Why it matters That obsessive self-checking pushed Fable to 16.4% on GDPx, but it's 2.6× the cost of Opus 5.5. At High effort it matches Opus 5.5 at Max for half the cost per task — meaning your effort-setting choice is now the single biggest lever on your inference bill. ⚡ Our take Anthropic is selling us more compute, not just more intelligence: token prices fall while bills climb. Anyone running Fable on Max by default is paying for the model's second-guessing.