benchgap
OpenAI · model

GPT-5.3 Codex benchmark scores

As of 2026-11-22, GPT-5.3 Codex (OpenAI) has measured scores on 19 benchmarks and estimated scores on 120 more.

BenchmarkScoreSource
AA-GPQA Diamond91.5%measured
AA-HLE42.5%measured
AA-SciCode48.6%estimated ± 3.6 pp, high confidence
Artificial Analysis Intelligence Index36.1%estimated ± 7.2 pp, medium confidence
CritPt16.9%measured
GDPval-AA36.4%estimated ± 6.2 pp, medium confidence
GPQA Diamond92.4%estimated ± 2.4 pp, high confidence
HLE48.2%estimated ± 8.5 pp, medium confidence
IFBench73.6%estimated ± 6.6 pp, medium confidence
AA-LCR83.3%measured
LiveCodeBench v689.7%estimated ± 2.2 pp, low confidence
AA-Omniscience Accuracy43.2%estimated ± 7.7 pp, medium confidence
SciCode51.3%estimated ± 2.9 pp, medium confidence
BrowseComp63.6%estimated ± 4.7 pp, high confidence
HLE w/ tools39.9%estimated ± 5.5 pp, medium confidence
IFEval93.0%estimated ± 2.8 pp, high confidence
LongBench v268.6%estimated ± 5.5 pp, medium confidence
τ²-bench results86.0%measured
AA Coding Index69.4%estimated ± 5.3 pp, medium confidence
AA-MMMU-Pro78.5%measured
APEX-Agents-AA31.6%estimated ± 3.6 pp, medium confidence
BioMysteryBench (human-difficult)47.2%estimated ± 6.8 pp, low confidence
DeepSearchQA75.4%estimated ± 7.9 pp, low confidence
SWE-bench Verified85.0%measured
BFCL v464.5%estimated ± 8.5 pp, low confidence
CyberGym76.3%estimated ± 11.0 pp, low confidence
JobBench33.7%measured
SWE-bench Pro56.8%measured
WideResearch77.7%estimated ± 1.2 pp, medium confidence
AA Agentic Index15.6%estimated ± 8.0 pp, medium confidence
MMLU-Pro85.1%estimated ± 7.0 pp, medium confidence
AA-IFBench75.4%measured
FailureSensorIQ67.6%estimated ± 1.8 pp, medium confidence
GPQA91.8%estimated ± 6.1 pp, medium confidence
MMLU88.2%estimated ± 8.3 pp, low confidence
Vals GPQA Diamond90.6%estimated ± 3.2 pp, high confidence
Vals LiveCodeBench87.3%measured
Vals MMLU-Pro87.7%estimated ± 3.0 pp, high confidence
MMMLU90.5%estimated ± 0.7 pp, low confidence
MMMU85.6%estimated ± 2.7 pp, high confidence
AA MMLU-Pro88.9%estimated ± 1.3 pp, medium confidence
Gert Labs52.0%measured
AA EnterpriseOps-Gym44.7%estimated ± 1.2 pp, medium confidence
ApprenticeBench4.7%estimated ± 6.0 pp, low confidence
ARC-AGI-187.4%estimated ± 7.8 pp, low confidence
ARC-AGI-263.1%estimated ± 14.4 pp, low confidence
CursorBench 3.152.5%estimated ± 5.6 pp, low confidence
CursorBench 3.253.5%estimated ± 3.6 pp, high confidence
FrontierCode 1.1 Main36.3%estimated ± 4.2 pp, medium confidence
FrontierSWE v210.1%estimated ± 8.3 pp, medium confidence
MazeBench0.9%estimated ± 2.7 pp, high confidence
OfficeQA Pro50.7%estimated ± 6.8 pp, medium confidence
OSWorld-Verified64.7%measured
ParseBench-163.4%estimated ± 6.2 pp, medium confidence
Terminal-Bench Hard55.3%estimated ± 4.0 pp, high confidence
Vals SWE-bench78.0%measured
Terminal-Bench 2.1 (Vals AI)68.8%estimated ± 6.4 pp, medium confidence
VulcanBench v385.2%estimated ± 5.6 pp, low confidence
CWE-Bench-AA52.9%estimated ± 8.1 pp, medium confidence
AA Cyber Index40.5%estimated ± 12.4 pp, low confidence
DeepsecBench-AA18.3%estimated ± 8.2 pp, medium confidence
HealthBench Professional (length-adjusted, Artificial Analysis)56.0%estimated ± 7.3 pp, low confidence
HealthBench Professional (raw, Artificial Analysis)63.3%estimated ± 0.8 pp, low confidence
AA ITBench45.2%estimated ± 4.6 pp, medium confidence
Bug Hunt Bench29.8%estimated ± 12.1 pp, low confidence
Chartography (tools)82.5%estimated ± 4.1 pp, low confidence
CursorBench 4.033.2%estimated ± 3.3 pp, medium confidence
CWE-bench v151.4%estimated ± 5.7 pp, medium confidence
DeepSWE56.3%estimated ± 7.2 pp, low confidence
FrontierCode 2.020.3%estimated ± 6.7 pp, low confidence
HLE w/o tools42.3%estimated ± 5.4 pp, medium confidence
OSWorld 2.00.0%estimated ± 3.6 pp, medium confidence
OSWorld 2.165.6%estimated ± 3.6 pp, high confidence
OSWorld 2.1 Offline60.4%estimated ± 10.1 pp, low confidence
PostTrainBench v1.114.2%estimated ± 1.3 pp, low confidence
Toolathlon-Verified73.3%estimated ± 3.0 pp, high confidence
VulcanBench Frontier v477.8%estimated ± 8.3 pp, low confidence
AA AutomationBench55.3%estimated ± 9.9 pp, low confidence
GDP.pdf16.4%estimated ± 4.1 pp, medium confidence
ARC-AGI-31.3%estimated ± 9.8 pp, medium confidence
ARC-AGI-3 Provider Adapter2.9%estimated ± 7.3 pp, medium confidence
CWE-bench48.2%estimated ± 6.2 pp, medium confidence
MKA36.4%estimated ± 4.6 pp, high confidence
SkillsBench official leaderboard (BenchFlow)52.8%estimated ± 9.5 pp, low confidence
BinaryAudit42.1%measured
OTelBench20.1%estimated ± 5.6 pp, medium confidence
Vibe Code Bench61.8%measured
Hermes Index33.8%estimated ± 5.7 pp, medium confidence
ParseBench-269.5%estimated ± 6.2 pp, medium confidence
CharXiv83.1%estimated ± 6.3 pp, low confidence
CharXiv w/o tools82.4%estimated ± 2.5 pp, high confidence
ExploitGym7.4%estimated ± 8.3 pp, low confidence
AI2D_TEST94.0%estimated ± 2.0 pp, medium confidence
C-Eval94.0%estimated ± 1.2 pp, low confidence
Claw-Eval63.0%estimated ± 4.5 pp, medium confidence
CountBench97.2%estimated ± 2.9 pp, medium confidence
DynaMath85.5%estimated ± 3.0 pp, medium confidence
ERQA68.3%estimated ± 3.1 pp, high confidence
MathVision89.2%estimated ± 3.1 pp, high confidence
MCP Atlas63.0%estimated ± 8.6 pp, medium confidence
MMLU-Redux95.4%estimated ± 3.3 pp, medium confidence
MMMU-Pro79.3%estimated ± 1.5 pp, high confidence
NL2Repo44.2%estimated ± 4.7 pp, high confidence
PostTrainBench v117.8%measured
RealWorldQA85.2%estimated ± 2.9 pp, medium confidence
ScreenSpot Pro76.6%estimated ± 10.8 pp, low confidence
Terminal-Bench 2.077.3%measured
Toolathlon45.3%estimated ± 7.4 pp, low confidence
V*100.0%estimated ± 7.5 pp, medium confidence
Video-MME (w/o subtitle)85.6%estimated ± 1.7 pp, medium confidence
Video-MME (with subtitle)87.3%estimated ± 3.7 pp, medium confidence
AA LiveCodeBench90.5%estimated ± 2.6 pp, medium confidence
HealthBench Hard31.8%estimated ± 8.6 pp, low confidence
LiveCodeBench Pro81.2%estimated ± 6.0 pp, low confidence
MedXpertQA (MM)76.3%estimated ± 6.5 pp, low confidence
React Native Evals81.9%estimated ± 2.7 pp, medium confidence
ResearchClawBench15.9%estimated ± 2.1 pp, medium confidence
SWE-Rebench58.2%measured
PostTrainBench15.4%estimated ± 1.2 pp, low confidence
Finance Agent v256.0%estimated ± 2.4 pp, medium confidence
BioMysteryBench (human-solvable)88.4%estimated ± 0.6 pp, medium confidence
FrontierCode 1.1 Extended54.9%estimated ± 3.0 pp, low confidence
HealthBench (raw)51.0%estimated ± 6.0 pp, medium confidence
HealthBench (length-adjusted)54.5%estimated ± 3.4 pp, medium confidence
HealthBench Professional58.2%estimated ± 4.5 pp, medium confidence
HealthBench Professional (raw)61.5%estimated ± 4.5 pp, high confidence
HLE-Verified38.7%estimated ± 2.0 pp, medium confidence
LABBench279.4%estimated ± 1.8 pp, low confidence
ScarfBench48.9%estimated ± 12.3 pp, low confidence
SimpleQA34.4%estimated ± 7.8 pp, low confidence
LiveCodeBench82.2%estimated ± 14.6 pp, low confidence
Agents' Last Exam31.9%estimated ± 9.7 pp, medium confidence
BabyVision w/ Python89.9%estimated ± 4.0 pp, low confidence
OpenHarmony Bench v1.053.7%estimated ± 3.2 pp, high confidence
MRCRv285.9%estimated ± 11.3 pp, low confidence
MMLongBench-Doc61.5%estimated ± 2.6 pp, low confidence
BioBasicsBench34.1%estimated ± 2.8 pp, medium confidence
LifeVal60.1%estimated ± 5.0 pp, medium confidence
MMMU-Pro w/ Python81.4%estimated ± 2.0 pp, high confidence