Home Model series GPT mini GPT-5.4 mini GPT-5.4 mini · MMMU-Pro · Standard/Vision 等格式未在所选材料中注明 · Scores and configurations GPT-5.4 mini · MMMU-Pro · Standard/Vision 等格式未在所选材料中注明 · Scores and configurations GPT-5.4 mini record for MMMU-Pro · Standard/Vision 等格式未在所选材料中注明. 2 score records for this release, including 2 exact values (may include historical records). Includes 2 source locations.
Scores and evaluation conditions GPT-5.4 mini · MMMU-Pro scores and configurations Metric and score Configuration and evaluation conditions Sources Score 78% Reported 78.0% · Unit % Published 2026-03-17
GPT-5.4 mini (xhigh) Settings The official comparison table labels GPT-5.4 mini (xhigh). Unknowns none recorded Tools Tools allowed Reasoning budget xhigh Framework not recorded Protocol MMMU-Pro with Python; exact format and tool harness are not stated. Conditions GPT-5.4 mini (xhigh); evaluation conducted in a research environment. View this score in rankings Score 76.6% Reported 76.6% · Unit % Published 2026-03-17
GPT-5.4 mini (xhigh) Settings The official comparison table labels GPT-5.4 mini (xhigh). Unknowns none recorded Tools Tools forbidden Reasoning budget xhigh Framework not recorded Protocol MMMU-Pro without tools; exact format is not stated. Conditions GPT-5.4 mini (xhigh); evaluation conducted in a research environment. View this score in rankings
Benchmark versions and metrics Score · % · Higher is better