OpenAI's Latest Model Scores 150 on Public IQ Test, Forcing Business Leaders to Recalculate

A new public benchmark result is cutting through a crowded news cycle focused on inflation and geopolitics. OpenAI’s GPT-5.4 Pro model has registered an IQ score of 150 on the TrackingAI leaderboard, which uses a Mensa Norway-style test. This marks a clear rise from the 136 score achieved by its predecessor, the o3 model, last year.

For executives, this isn't just an academic milestone. The jump translates a complex surge in machine reasoning into a single, digestible figure. It arrives as OpenAI promotes GPT-5.4 Pro for its enhanced coding, tool use, and ability to handle professional workloads. The company also reports the model sets new standards on other technical benchmarks like GDPval and OSWorld-Verified.

The consistent upward trend across different measurement systems is what commands attention. While IQ-style tests are imperfect—they capture a narrow slice of cognitive skill and can be influenced by test design—a cluster of advancing scores is harder to ignore. It signals progress in pattern recognition and multi-step problem solving that directly applies to automating document workflows, customer support, code generation, and internal operations.

This development inserts a new variable into strategic planning. The primary question for businesses is no longer if AI capabilities are improving, but how rapidly those improvements should alter budgets for software, cloud infrastructure, and staffing. A model that tests at this level while gaining proficiency in real-world computer use suggests a shift from experimentation to tangible deployment is accelerating.

OpenAI’s position atop a public leaderboard, ahead of rivals like Claude and Gemini, also serves a competitive purpose. It provides a straightforward signal to developers and enterprise buyers in a market where differentiating between models is increasingly difficult.

As markets parse economic data this week, this benchmark serves as a reminder: the growth curve of machine intelligence itself is becoming a factor in investment and operational decisions, with implications that stretch far beyond a test score.

Source: Crypto Slate