OpenAI's GPT-5.4 Pro Achieves 150 IQ Score, Surpassing 99.96% of Humans on MESNA Norway Test

OpenAI's GPT-5.4 Pro model scored 150 on the MESNA Norway standardized IQ test, surpassing 99.96% of all humans and breaking OpenAI's own previous record. The achievement appears on TrackingAI's public leaderboard for AI model rankings and represents capability gains that are beginning to outpace typical product-cycle developments. The result provides markets with a fresh signal about the accelerating pace of AI advancement, marking what observers consider a significant milestone in artificial intelligence capabilities.

GPT-5.4 Pro's IQ Test Performance and Benchmark Results

• ▪OpenAI's GPT-5.4 Pro model was tested on the MESNA Norway test.
• ▪OpenAI's GPT-5.4 Pro model achieved an IQ score higher than 99.96% of all human beings.
• ▪OpenAI's GPT-5.4 Pro model broke OpenAI's own previous record.
• ▪TrackingAI's public leaderboard now places OpenAI's GPT-5.4 Pro model on its rankings.
• ▪OpenAI's GPT-5.4 Pro model achieved an IQ score of 150.

AI Capability Advancement and Market Implications

• ▪OpenAI's GPT-5.4 Pro model's IQ achievement gave markets a fresh signal about AI capability gains.
• ▪AI capability gains are starting to outpace the usual product-cycle noise.

Perspective of AI safety researchers

• ▪The MESNA Norway test result provides markets with a fresh signal about the pace of AI capability advancement.
• ▪GPT-5.4 Pro's 150 IQ score represents AI capability gains that are outpacing the usual product-cycle noise.

Perspective of OpenAI

• ▪GPT-5.4 Pro's performance on the MESNA Norway test marks a significant milestone in AI capability advancement.

1 source