GPT-5.4 | Humanity's Last Exam Full set — No tools
OpenAI's GPT-5.4 evaluation results on Humanity's Last Exam Full set — No tools
Score: 39.8%
Reasoning level: xhigh
OpenAI's GPT-5.4 evaluation results on Humanity's Last Exam Full set — No tools
Score: 39.8%
Reasoning level: xhigh