benchgap
Anthropic · model

Claude 4.1 Opus benchmark scores

As of 2026-11-22, Claude 4.1 Opus (Anthropic) has measured scores on 3 benchmarks and estimated scores on 55 more.

BenchmarkScoreSource
AA-GPQA Diamond82.2%estimated ± 12.1 pp, low confidence
AA-HLE22.4%estimated ± 9.4 pp, medium confidence
AA-SciCode45.9%estimated ± 3.8 pp, high confidence
Artificial Analysis Intelligence Index18.6%measured
CritPt3.4%estimated ± 4.7 pp, high confidence
GDPval-AA28.5%estimated ± 6.2 pp, medium confidence
GPQA Diamond89.1%estimated ± 2.6 pp, high confidence
HLE32.7%estimated ± 12.0 pp, low confidence
AA-Omniscience Accuracy28.3%estimated ± 7.3 pp, medium confidence
SciCode45.1%estimated ± 3.6 pp, high confidence
BrowseComp69.6%estimated ± 6.2 pp, medium confidence
AA Coding Index49.3%estimated ± 5.3 pp, medium confidence
APEX-Agents-AA24.6%estimated ± 3.6 pp, medium confidence
BioMysteryBench (human-difficult)41.6%estimated ± 8.3 pp, low confidence
DeepSearchQA77.1%estimated ± 8.2 pp, low confidence
SWE-bench Verified74.5%measured
CyberGym47.6%estimated ± 11.4 pp, low confidence
JobBench21.9%measured
SWE-bench Pro52.4%estimated ± 4.5 pp, high confidence
WideResearch75.1%estimated ± 1.2 pp, medium confidence
MMLU-Pro85.7%estimated ± 5.4 pp, medium confidence
GPQA82.9%estimated ± 7.8 pp, medium confidence
MMLU89.9%estimated ± 3.8 pp, medium confidence
Vals GPQA Diamond82.4%estimated ± 5.7 pp, medium confidence
Vals LiveCodeBench75.8%estimated ± 10.0 pp, low confidence
Vals MMLU-Pro83.7%estimated ± 4.2 pp, high confidence
MMMLU86.8%estimated ± 3.0 pp, medium confidence
AA MMLU-Pro85.9%estimated ± 1.2 pp, high confidence
Gert Labs33.7%estimated ± 7.7 pp, medium confidence
AA EnterpriseOps-Gym42.5%estimated ± 1.2 pp, low confidence
ApprenticeBench0.4%estimated ± 6.0 pp, low confidence
ARC-AGI-178.0%estimated ± 7.9 pp, low confidence
ARC-AGI-239.7%estimated ± 14.6 pp, low confidence
FrontierCode 1.1 Main17.8%estimated ± 4.2 pp, low confidence
MazeBench0.0%estimated ± 5.1 pp, medium confidence
OSWorld-Verified61.5%estimated ± 8.4 pp, low confidence
Vals SWE-bench68.8%estimated ± 6.6 pp, medium confidence
HealthBench Professional (length-adjusted, Artificial Analysis)51.5%estimated ± 4.3 pp, low confidence
HealthBench Professional (raw, Artificial Analysis)54.0%estimated ± 1.1 pp, low confidence
DeepSWE37.5%estimated ± 12.7 pp, low confidence
HLE w/o tools30.7%estimated ± 4.9 pp, high confidence
GDP.pdf16.4%estimated ± 4.1 pp, low confidence
Vibe Code Bench1.6%estimated ± 8.4 pp, medium confidence
C-Eval91.1%estimated ± 1.0 pp, medium confidence
Claw-Eval57.0%estimated ± 4.5 pp, medium confidence
MCP Atlas47.8%estimated ± 8.6 pp, medium confidence
MMLU-Redux93.8%estimated ± 1.2 pp, medium confidence
NL2Repo34.5%estimated ± 6.2 pp, medium confidence
PostTrainBench v113.5%estimated ± 5.5 pp, low confidence
SuperGPQA61.7%estimated ± 10.1 pp, low confidence
Toolathlon33.1%estimated ± 7.4 pp, low confidence
React Native Evals77.2%estimated ± 2.7 pp, medium confidence
ResearchClawBench15.7%estimated ± 2.6 pp, low confidence
SWE-Rebench57.1%estimated ± 3.6 pp, medium confidence
HealthBench (length-adjusted)54.1%estimated ± 3.7 pp, medium confidence
HealthBench Professional56.7%estimated ± 4.2 pp, low confidence
LiveCodeBench75.9%estimated ± 14.6 pp, low confidence
BioBasicsBench22.2%estimated ± 10.2 pp, low confidence