AI Models Are Now Just Expensive Bragging Rights

Claude Fable 5.2 just scored 65 on the Artificial Analysis Intelligence Index, which is apparently a big deal because the median score is only 40. This model can handle text and images, has a massive one-million-token memory window, and managed to generate a whopping 275 million tokens during testing—nearly four times the typical amount. All this verbosity and intelligence comes with a hefty price tag: $8 per million input tokens and $40 per million output tokens, which is four times more expensive than average models. For developers, this means another round of checking whether paying four times more for a slightly smarter model is actually worth it, especially when the main achievement seems to be scoring higher on a benchmark few people understand. The takeaway? The AI industry has perfected the art of selling expensive digital report cards while convincing everyone it’s progress.