Home Model series GPT unsuffixed models GPT-5.4 GPT-5.4 · Humanity's Last Exam · 2,500-question benchmark · Scores and configurations GPT-5.4 · Humanity's Last Exam · 2,500-question benchmark · Scores and configurations GPT-5.4 record for Humanity's Last Exam · 2,500-question benchmark. 4 score records for this release, including 4 exact values (may include historical records). Includes 2 source locations.
Scores and evaluation conditions GPT-5.4 · Humanity's Last Exam scores and configurations Metric and score Configuration and evaluation conditions Sources Accuracy 39.8% Reported 39.8% · Unit % Published Unconfirmed
GPT-5.4 (xhigh) Settings The release article states evaluations use xhigh unless otherwise noted. Unknowns none recorded Tools Tools forbidden Reasoning budget xhigh Framework not recorded Protocol Humanity's Last Exam; no tools. Conditions Official release table labels GPT-5.4 (xhigh); the article says evaluations use xhigh unless otherwise noted and were run in a research environment. View this score in rankings Accuracy 42.7% Reported 42.7% · Unit % Published Unconfirmed
GPT-5.4 Pro Settings Pro model variant; the article reports Pro separately from GPT-5.4. Unknowns The Pro column is not a reasoning-effort label; the exact effort for each Pro evaluation is not separately stated. Tools Tools forbidden Reasoning budget not recorded Framework not recorded Protocol Humanity's Last Exam; no tools. Conditions Official release table labels GPT-5.4 Pro; research evaluation environment. View this score in rankings Accuracy 52.1% Reported 52.1% · Unit % Published Unconfirmed
GPT-5.4 (xhigh) Settings The release article states evaluations use xhigh unless otherwise noted. Unknowns none recorded Tools Tools allowed Reasoning budget xhigh Framework not recorded Protocol Humanity's Last Exam; with tools. Conditions Official release table labels GPT-5.4 (xhigh); the article says evaluations use xhigh unless otherwise noted and were run in a research environment. View this score in rankings Accuracy 58.7% Reported 58.7% · Unit % Published Unconfirmed
GPT-5.4 Pro Settings Pro model variant; the article reports Pro separately from GPT-5.4. Unknowns The Pro column is not a reasoning-effort label; the exact effort for each Pro evaluation is not separately stated. Tools Tools allowed Reasoning budget not recorded Framework not recorded Protocol Humanity's Last Exam; with tools. Conditions Official release table labels GPT-5.4 Pro; research evaluation environment. View this score in rankings
Benchmark versions and metrics Aliases HLE、Humanity’s Last Exam、Humanity's Last Exam
Accuracy · % · Higher is better