benchgap
OpenAI · model

o3 benchmark scores

As of 2026-11-21, o3 (OpenAI) has measured scores on 16 benchmarks and estimated scores on 97 more.

BenchmarkScoreSource
AA-GPQA Diamond82.7%measured
AA-HLE20.1%measured
AA-SciCode51.8%estimated ± 5.1 pp, low confidence
AIME2695.2%estimated ± 1.4 pp, high confidence
Artificial Analysis Intelligence Index18.4%estimated ± 3.6 pp, high confidence
CritPt1.1%measured
GDPval-AA13.8%estimated ± 11.0 pp, low confidence
GPQA Diamond87.5%estimated ± 1.8 pp, high confidence
HLE23.4%estimated ± 8.5 pp, medium confidence
IFBench69.4%estimated ± 6.6 pp, medium confidence
AA-LCR74.7%measured
LiveCodeBench v688.2%estimated ± 2.2 pp, medium confidence
AA-Omniscience Accuracy29.4%estimated ± 4.8 pp, high confidence
SciCode44.5%estimated ± 2.9 pp, low confidence
BrowseComp67.1%estimated ± 13.0 pp, low confidence
HLE w/ tools32.9%estimated ± 5.5 pp, low confidence
IFEval92.9%estimated ± 2.8 pp, high confidence
LongBench v261.5%estimated ± 5.5 pp, medium confidence
τ²-bench results80.7%measured
AA Coding Index63.1%estimated ± 11.5 pp, low confidence
AA-MMMU-Pro70.1%measured
APEX-Agents-AA1.6%estimated ± 11.4 pp, low confidence
BioMysteryBench (human-difficult)18.0%estimated ± 6.8 pp, low confidence
SWE-bench Verified81.4%estimated ± 13.1 pp, low confidence
BFCL v461.8%estimated ± 8.5 pp, low confidence
SWE-bench Pro57.2%estimated ± 4.7 pp, high confidence
AA Agentic Index19.7%estimated ± 12.7 pp, low confidence
MMLU-Pro85.2%estimated ± 4.8 pp, high confidence
AA-IFBench71.4%measured
FailureSensorIQ67.6%measured
GPQA86.0%estimated ± 4.5 pp, high confidence
MMLU93.1%estimated ± 1.2 pp, medium confidence
Vals GPQA Diamond84.1%measured
Vals LiveCodeBench83.9%measured
Vals MMLU-Pro85.6%measured
MMMLU86.5%estimated ± 0.7 pp, medium confidence
MMMU80.2%estimated ± 2.7 pp, high confidence
Chess Puzzles36.0%measured
AA MMLU-Pro85.3%measured
ARC-AGI-176.3%estimated ± 6.5 pp, medium confidence
ARC-AGI-233.5%estimated ± 13.8 pp, low confidence
CursorBench 3.259.1%estimated ± 7.2 pp, low confidence
EBR-bench18.9%estimated ± 10.4 pp, low confidence
FrontierCode 1.1 Main38.0%estimated ± 6.2 pp, medium confidence
FrontierMath v2 (Tier 4)4.8%estimated ± 4.0 pp, high confidence
FrontierMath v2 (Tiers 1-3)33.3%measured
Furniture Assembly33.5%estimated ± 9.3 pp, medium confidence
MaxxBench47.8%estimated ± 8.8 pp, medium confidence
Mystery Game Puzzles29.0%measured
NYT Connections79.5%estimated ± 12.6 pp, low confidence
OfficeQA Pro50.3%estimated ± 6.8 pp, low confidence
ParseBench-153.9%estimated ± 6.2 pp, medium confidence
Vals SWE-bench52.9%estimated ± 9.0 pp, medium confidence
VulcanBench v382.9%estimated ± 5.6 pp, low confidence
HealthBench Professional (raw, Artificial Analysis)5.5%estimated ± 2.1 pp, low confidence
Chartography (tools)77.0%estimated ± 4.1 pp, low confidence
CursorBench 4.040.7%estimated ± 4.5 pp, high confidence
FrontierPhysics2.7%estimated ± 5.0 pp, medium confidence
HLE w/o tools32.3%estimated ± 4.9 pp, high confidence
SpireClimb0.1%estimated ± 9.3 pp, medium confidence
StealthMark43.6%estimated ± 9.2 pp, medium confidence
ARC-AGI-30.0%estimated ± 8.6 pp, low confidence
ARC-AGI-3 Provider Adapter2.7%estimated ± 7.3 pp, low confidence
SkillsBench official leaderboard (BenchFlow)55.3%estimated ± 6.2 pp, medium confidence
BinaryAudit35.0%estimated ± 8.9 pp, low confidence
CompileBench100.0%estimated ± 10.0 pp, low confidence
OTelBench31.2%estimated ± 4.1 pp, low confidence
Vibe Code Bench0.0%estimated ± 6.6 pp, low confidence
FACTS Multimodal42.1%estimated ± 5.6 pp, medium confidence
ParseBench-258.1%estimated ± 6.2 pp, medium confidence
CharXiv77.7%estimated ± 6.3 pp, low confidence
CharXiv w/o tools77.0%estimated ± 2.5 pp, medium confidence
ExploitBench49.1%estimated ± 7.2 pp, medium confidence
AI2D_TEST88.1%estimated ± 2.0 pp, medium confidence
C-Eval91.9%estimated ± 1.1 pp, medium confidence
CountBench91.6%estimated ± 2.9 pp, low confidence
DynaMath71.6%estimated ± 3.0 pp, low confidence
ERQA43.9%estimated ± 3.1 pp, medium confidence
HMMT Feb 202686.2%estimated ± 3.1 pp, high confidence
HMMT Nov 202592.6%estimated ± 1.5 pp, medium confidence
MathVision78.1%estimated ± 3.1 pp, high confidence
MMAnswerBench82.3%estimated ± 2.4 pp, medium confidence
MMLU-Redux93.6%estimated ± 3.3 pp, high confidence
MMMU-Pro71.5%estimated ± 1.5 pp, high confidence
NL2Repo45.7%estimated ± 7.9 pp, low confidence
RealWorldQA79.8%estimated ± 2.9 pp, low confidence
ScreenSpot Pro71.3%estimated ± 10.8 pp, low confidence
V*51.0%estimated ± 7.5 pp, low confidence
Video-MME (w/o subtitle)79.8%estimated ± 1.7 pp, medium confidence
Video-MME (with subtitle)81.8%estimated ± 3.7 pp, low confidence
AA LiveCodeBench80.8%measured
HealthBench Hard10.4%estimated ± 8.6 pp, low confidence
MedXpertQA (MM)51.5%estimated ± 6.5 pp, low confidence
MMLU-Pro (Arcee)81.3%estimated ± 4.0 pp, medium confidence
DAYJOB1.3%estimated ± 3.2 pp, high confidence
DAYJOB Healthcare0.7%estimated ± 4.2 pp, high confidence
BioMysteryBench (human-solvable)81.7%estimated ± 0.6 pp, low confidence
HealthBench (raw)27.8%estimated ± 6.0 pp, low confidence
HealthBench (length-adjusted)46.0%estimated ± 3.4 pp, low confidence
HealthBench Professional58.1%estimated ± 4.5 pp, low confidence
HealthBench Professional (raw)39.9%estimated ± 4.5 pp, medium confidence
HLE-Verified0.2%estimated ± 4.6 pp, low confidence
LABBench280.7%estimated ± 1.8 pp, low confidence
BlenderWorld (text-only)70.7%estimated ± 2.8 pp, medium confidence
Cookie-Bench78.8%estimated ± 1.3 pp, medium confidence
SimpleQA30.0%estimated ± 7.8 pp, low confidence
BabyVision w/ Python84.3%estimated ± 4.0 pp, low confidence
SEC-Bench Pro48.5%estimated ± 9.3 pp, medium confidence
MRCRv283.9%estimated ± 11.3 pp, low confidence
MMLongBench-Doc61.2%estimated ± 2.6 pp, low confidence
BioBasicsBench33.9%estimated ± 10.7 pp, low confidence
MMMU-Pro w/ Python73.8%estimated ± 2.0 pp, high confidence
ExploitBench Internal Port0.2%estimated ± 1.4 pp, medium confidence