Sources: 
The latest rankings of AI intelligence, based on the Mensa Norway IQ test, reveal that text-only AI models dominate the leaderboard.
OpenAI’s text-only o3 model leads with a score of
135, placing it firmly in the
"genius" category, surpassing the average human IQ range of 90 to 110.
This Mensa test is a highly respected and challenging exam designed to evaluate human intelligence, making these AI scores particularly notable.
Surprisingly, all of the top 10 smartest AI models are text-only systems, which means they do not have the ability to read or process images. This dominance suggests that current AI advancements in natural language processing have outpaced multimodal capabilities in IQ test performance.
The results highlight the strength of language-based AI in reasoning and problem-solving tasks traditionally used to measure intelligence.
"Topping the chart is OpenAI’s text-only o3 model, scoring a 135 on the Mensa IQ test, which puts it in the ‘genius’ category," the data notes.
The findings challenge assumptions that multimodal AI, which can interpret images and text, would outperform text-only systems in such cognitive benchmarks.
As AI development continues, these rankings provide insight into which architectures currently excel in human-like intelligence assessments.
Sources: OpenAI’s text-only o3 AI model tops the Mensa Norway IQ test with a score of 135, placing it in the genius category. Surprisingly, all top 10 smartest AI models are text-only systems, highlighting their dominance despite lacking image processing capabilities.
