AI Model Scores Mensa Norway IQ 135, Surpassing Average Human Reasoning
The latest round of benchmark testing has put a spotlight on a new language model that cracked the Mensa Norway IQ exam with a score of 135. That figure sits well above the average human range, suggesting the system can handle logic puzzles and pattern‑recognition tasks that traditionally separate high‑scoring test‑takers from the broader population. What’s notable is that the model isn’t a visual‑heavy version; its text‑only sibling, GPT‑5.2 Pro Vision, posted the same 135, indicating the achievement isn’t tied to image processing capabilities.
In a field where scores often hover near baseline human performance, breaking the 130‑plus threshold raises questions about how far pure reasoning can be pushed by current architectures. The result also feeds into a broader conversation about what “intelligence” means when an algorithm can out‑score most people on a standardized test designed for human cognition. As the data rolls in, analysts are watching to see whether this marks a meaningful step forward or simply a narrow victory in a specific testing context.
While it may not top the chart, its performance confirms that it operates far beyond human-average reasoning levels when evaluated purely on logic and pattern-based intelligence. Mensa Norway IQ: 135 Not far behind its text-only counterpart is GPT-5.2 Pro Vision, scoring 135 on the Mensa Norway test. This still places it firmly within the range of very high human intelligence.
This is well above both the global average and the typical threshold associated with advanced reasoning ability. Note that this score comes from a vision-enabled model - an AI model that can process and reason over visual information (like input images), and not just text.
The scores speak for themselves: Gemini 3 and GPT‑5.2 Pro Vision each hit 135 on the Mensa Norway IQ test, well above the human average of 100. Yet the numbers tell only part of the story. While the models “operate far beyond human‑average reasoning levels when evaluated purely on logic and pattern‑based intelligence,” they lack consciousness, emotions and lived experience.
Does a higher IQ score equate to overall superiority? Unclear. The tests were built for human cognition, not for machines that process text without awareness.
Moreover, the article notes that the models “may not top the chart,” suggesting that other, perhaps more specialized, systems could perform differently. What remains uncertain is whether these results translate into broader, real‑world understanding or merely reflect pattern‑matching prowess. The data, however, confirms that at least on this particular metric, the latest large language models exceed typical human performance.
Whether that constitutes genuine intelligence or simply a different kind of computational ability is still an open question.
Further Reading
- OpenAI’s o3 scores 136 on Mensa Norway test, surpassing 98% of human population - CryptoSlate
- Ranked: The Smartest AI Models, by IQ - Visual Capitalist
- Comparing the IQ of AI Models - Voronoi
- The Ultimate AI Model Guide 2025: Best Picks for Every Task - Pacifyca
More in LLMs & Generative AI
OpenAI launches App Directory, accepts ChatGPT apps with privacy notices
OpenAI has opened the door for external creators to plug their tools straight into ChatGPT, rolling...
Google pushes Gemini rollout to 2026, delays Android Assistant swap
Why does this matter now? For years, Android users have expected a smooth handoff from the familiar...
OpenAI now adds tone controls and streamlined email editing to ChatGPT
OpenAI just rolled out a set of tweaks that feel oddly personal for a chatbot. While the core model...
Google releases FunctionGemma, a tiny model for natural‑language mobile control
Here's the thing: mobile phones are the most common computing platform, yet most AI assistants...
Chinese scammers used AI images for refund fraud; police held buyer eight days
When a shopper in Shanghai posted a video of a defective product, the footage looked...
Common Questions Answered
What score did the new language model achieve on the Mensa Norway IQ test?
The model earned a Mensa Norway IQ score of 135, which is well above the human average of 100. This places the AI in the very high intelligence range typically reserved for top human test‑takers.
How does GPT-5.2 Pro Vision's performance compare to its text‑only sibling on the Mensa Norway IQ exam?
GPT-5.2 Pro Vision matched its text‑only counterpart by also scoring 135 on the Mensa Norway IQ test. The identical results demonstrate that both the visual and text‑only versions exhibit the same level of logic and pattern‑recognition ability.
Which AI models are reported to have reached a 135 score on the Mensa Norway IQ test?
The article states that both Gemini 3 and GPT‑5.2 Pro Vision each achieved a score of 135. These results put the two models firmly within the range of very high human intelligence on this specific benchmark.
What limitations does the article highlight regarding AI models achieving high IQ scores?
Despite their high IQ scores, the models lack consciousness, emotions, and lived experience, which are essential aspects of human cognition. The article also notes that the Mensa Norway test was designed for humans, so its relevance to machine intelligence remains uncertain.