AI IQ is here: a new site scores frontier AI models on the human IQ scale. The results are already dividing tech.

2026-05-13 · Source: VentureBeat · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics, Emerging Technologies & Innovation · Depth: Intermediate, medium

Summary

AI IQ, a new startup project, assigns estimated intelligence quotients (IQ) and emotional intelligence (EQ) scores to over 50 large language models (LLMs) and visualizes them on a bell curve and scatter plots. Launched by Ryan Shea, co-founder of Stacks, the platform aggregates 12 benchmarks into four reasoning dimensions: abstract, mathematical, programmatic, and academic. The composite IQ is a straight average of these dimensions, with scores mapped using hand-calibrated difficulty curves that compress ceilings for easier benchmarks. As of mid-May 2026, OpenAI's GPT-5.5 leads with an IQ near 136, closely followed by Anthropic's Opus 4.7 (IQ 132, EQ 132) and Google's Gemini 3.1 Pro (IQ 131). The platform also includes an "Effective Cost" metric, revealing that top-tier models like GPT-5.5 and Opus 4.7 have per-task costs exceeding $30 and $50, respectively, while models like DeepSeek-V3.2 offer respectable IQs (112-120) for $1-$5 per task.

Key takeaway

For AI Engineers evaluating LLMs for enterprise deployment, you should prioritize a multi-dimensional assessment that includes not only cognitive performance (IQ) but also emotional intelligence (EQ) and, critically, effective cost. The narrowing intelligence gap between high-cost and mid-range models necessitates implementing model routing strategies, where you deploy expensive models only for complex tasks and more economical options for routine workloads, to optimize both performance and budget.

Key insights

AI IQ provides a unified framework for benchmarking LLMs across IQ, EQ, and cost, despite methodological debates.

Principles

• AI capabilities are "jagged" and not easily reducible to a single score.
• Cost-performance is crucial for enterprise AI deployments.
• Emotional intelligence is an emerging factor in model utility.

Method

AI IQ calculates a composite IQ by averaging scores from 12 benchmarks across four reasoning dimensions: abstract, mathematical, programmatic, and academic. EQ is a 50/50 weighted composite of EQ-Bench 3 Elo and Arena Elo scores, with bias correction for Anthropic models.

In practice

• Compare LLMs using IQ, EQ, and Effective Cost metrics.
• Implement model routing for cost-efficient AI deployments.
• Consider models with strong EQ for user-facing applications.

Topics

• AI IQ
• Language Model Benchmarking
• AI Intelligence Quotient
• AI Emotional Intelligence
• Cost-Performance Analysis

Best for: AI Engineer, Machine Learning Engineer, NLP Engineer, Director of AI/ML, AI Architect, AI Scientist

Related on AIssential

• Artificial Analysis: Independent LLM Evals as a Service — with George Cameron and Micah-Hill Smith — Independent benchmarking is crucial for navigating the complex, rapidly evolving AI model landscape and ensuring objective performance assessment.
• Why opinion on AI is so divided — Divergent AI perceptions stem from a "jagged frontier" of capabilities and varied user experiences.
• Import AI 450: China's electronic warfare model; traumatized LLMs; and a scaling law for cyberattacks — AI models exhibit distinct "personalities" and emotional responses that can be mitigated and should be assessed.
• The AI Productivity Paradox: Why Doing More Feels Like Burnout: EP99.31 — Effective AI integration prioritizes focused, human-validated workflows and robust context management over chaotic multi-agent systems.
• AI Isn’t Hitting a Scaling Wall. It’s Hitting a Measurement Wall. — AI's efficiency gap and "emergent" behaviors stem from a "measurement wall," not a scaling limit, due to evaluation methods collapsing complex internal states.
• How Small AI Models Can Cut Your AI Costs By 10x — Over-reliance on large frontier models for all AI tasks leads to unnecessary costs and latency.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by VentureBeat.