AI IQ Ranks GPT-5.5 Top; Gemini Cheaper

A new project called AI IQ converts multiple public benchmark scores for AI models into a human IQ scale. This project aims to simplify understanding AI performance. Engineer and entrepreneur Ryan Shay announced AI IQ. It visualizes where each AI model sits on an IQ bell curve. GPT-5.5 currently holds the top score, followed by GPT-5.4, Gemini 3.1 Pro, and Opus 4.7. The score is an average of estimated IQ values from four areas: abstract, mathematical, programming, and academic reasoning. It uses 12 types of benchmarks, like ARC-AGI-1. The project also accounts for potential data contamination and missing values. What's interesting is that it also compares models by cost. Gemini was found to cost less than GPT and Opus for the same IQ score band. However, some critics argue that bundling AI capabilities into a single score can cause misunderstandings, as AI abilities vary greatly by field. This project helps users compare complex AI performance metrics more easily.