benchgap
OpenAI · model

o3-pro benchmark scores

As of 2026-11-21, o3-pro (OpenAI) has measured scores on 2 benchmarks and estimated scores on 25 more.

BenchmarkScoreSource
AA-GPQA Diamond84.5%measured
AA-HLE24.0%estimated ± 2.7 pp, high confidence
Artificial Analysis Intelligence Index21.9%measured
CritPt1.8%estimated ± 3.8 pp, high confidence
GPQA Diamond89.9%estimated ± 2.6 pp, high confidence
HLE29.6%estimated ± 10.1 pp, low confidence
AA-Omniscience Accuracy31.1%estimated ± 7.3 pp, medium confidence
BioMysteryBench (human-difficult)43.1%estimated ± 8.3 pp, low confidence
MMLU-Pro86.1%estimated ± 5.4 pp, medium confidence
GPQA85.5%estimated ± 6.1 pp, medium confidence
MMLU89.9%estimated ± 3.8 pp, medium confidence
Vals GPQA Diamond83.6%estimated ± 3.2 pp, high confidence
Vals MMLU-Pro85.4%estimated ± 3.0 pp, high confidence
MMMLU88.1%estimated ± 1.7 pp, high confidence
AA MMLU-Pro86.7%estimated ± 1.2 pp, high confidence
ARC-AGI-179.8%estimated ± 7.9 pp, low confidence
ARC-AGI-245.2%estimated ± 14.6 pp, low confidence
HealthBench Professional (length-adjusted, Artificial Analysis)52.5%estimated ± 4.3 pp, low confidence
HealthBench Professional (raw, Artificial Analysis)57.0%estimated ± 1.1 pp, low confidence
HLE w/o tools33.4%estimated ± 4.9 pp, high confidence
C-Eval92.1%estimated ± 1.0 pp, medium confidence
MMLU-Redux94.4%estimated ± 1.2 pp, medium confidence
SuperGPQA62.5%estimated ± 10.1 pp, low confidence
HealthBench (length-adjusted)54.1%estimated ± 3.7 pp, medium confidence
HealthBench Professional56.7%estimated ± 4.2 pp, low confidence
SimpleQA30.9%estimated ± 7.8 pp, low confidence
BioBasicsBench25.6%estimated ± 10.2 pp, low confidence