benchgap
Anthropic · model

Claude Sonnet 5 benchmark scores

As of 2026-11-22, Claude Sonnet 5 (Anthropic) has measured scores on 53 benchmarks and estimated scores on 146 more.

BenchmarkScoreSource
AA-GPQA Diamond91.1%measured
AA-HLE41.3%measured
AA-SciCode54.3%measured
AIME2699.2%estimated ± 1.2 pp, medium confidence
Artificial Analysis Intelligence Index38.2%measured
CritPt16.9%measured
GDPval-AA48.2%measured
GPQA Diamond90.3%estimated ± 1.8 pp, high confidence
HLE57.4%measured
AA-LCR82.0%measured
LiveCodeBench v687.5%estimated ± 2.2 pp, medium confidence
AA-Omniscience Accuracy40.1%measured
SciCode41.5%estimated ± 2.9 pp, low confidence
Terminal-Bench 2.180.4%measured
BrowseComp84.7%measured
HLE w/ tools57.4%measured
LongBench v267.4%estimated ± 5.5 pp, medium confidence
VITA-Bench46.1%estimated ± 6.7 pp, low confidence
τ²-bench results98.1%estimated ± 3.1 pp, high confidence
AA Coding Index71.6%measured
AA-MMMU-Pro77.3%measured
APEX-Agents40.8%estimated ± 9.4 pp, low confidence
APEX-Agents-AA38.0%estimated ± 3.5 pp, low confidence
BioMysteryBench (human-difficult)47.2%estimated ± 6.8 pp, low confidence
DeepSearchQA92.5%estimated ± 4.5 pp, high confidence
SWE-bench Verified85.2%measured
AutomationBench23.5%estimated ± 7.1 pp, low confidence
BFCL v476.5%estimated ± 2.7 pp, medium confidence
CyberGym79.5%estimated ± 3.1 pp, low confidence
JobBench51.1%estimated ± 8.7 pp, medium confidence
SkillsBench56.7%estimated ± 4.1 pp, medium confidence
SWE-bench Pro63.2%measured
WideResearch81.4%estimated ± 2.1 pp, medium confidence
AA Agentic Index44.3%measured
MMLU-Pro85.7%estimated ± 2.5 pp, low confidence
FailureSensorIQ66.9%estimated ± 1.7 pp, high confidence
GPQA92.2%estimated ± 1.9 pp, high confidence
MMLU90.5%estimated ± 0.9 pp, medium confidence
Vals GPQA Diamond88.9%measured
Vals LiveCodeBench82.4%measured
Vals MMLU-Pro87.5%measured
MMMLU90.5%estimated ± 0.7 pp, low confidence
MMMU84.8%estimated ± 2.7 pp, high confidence
Chess Puzzles45.0%measured
AA MMLU-Pro87.4%estimated ± 0.7 pp, high confidence
Gert Labs47.0%measured
AA-AnalystAgent46.3%measured
AA EnterpriseOps-Gym44.7%measured
AA Harvey LAB v1.092.9%estimated ± 1.2 pp, medium confidence
AA Tau3 Banking37.3%measured
AA Terminal-Bench 2.180.5%measured
ApprenticeBench16.0%measured
ARC-AGI-195.7%estimated ± 0.8 pp, low confidence
ARC-AGI-273.6%estimated ± 13.0 pp, low confidence
BabaIsBench67.0%measured
CursorBench 3.154.6%estimated ± 3.4 pp, medium confidence
CursorBench 3.261.5%measured
EBR-bench25.6%estimated ± 10.4 pp, low confidence
Terminal-Bench 3.014.6%measured
FrontierCode 1.1 Main42.7%measured
FrontierMath v2 (Tier 4)29.3%measured
FrontierMath v2 (Tiers 1-3)65.6%measured
FrontierSWE v213.6%estimated ± 5.8 pp, low confidence
Furniture Assembly42.0%estimated ± 7.1 pp, medium confidence
MaxxBench53.8%estimated ± 5.6 pp, medium confidence
MazeBench0.0%measured
MirrorCode63.3%measured
Mystery Game Puzzles35.0%measured
NYT Connections75.1%measured
OfficeQA Pro65.8%estimated ± 2.5 pp, low confidence
OSWorld-Verified81.2%measured
ParseBench-161.2%measured
Terminal-Bench 4.112.1%measured
Terminal-Bench Hard55.8%estimated ± 2.8 pp, medium confidence
Vals SWE-bench79.6%measured
Terminal-Bench 2.1 (Vals AI)74.5%measured
VulcanBench v382.1%estimated ± 5.1 pp, low confidence
CWE-Bench-AA57.8%estimated ± 5.2 pp, medium confidence
AA Cyber Index26.5%estimated ± 9.2 pp, low confidence
DeepsecBench-AA16.0%estimated ± 8.2 pp, medium confidence
Harvey LAB-AA v1.17.4%estimated ± 4.0 pp, low confidence
HealthBench Professional (length-adjusted, Artificial Analysis)57.5%estimated ± 4.3 pp, medium confidence
HealthBench Professional (raw, Artificial Analysis)63.3%estimated ± 0.8 pp, low confidence
AA ITBench43.9%estimated ± 3.2 pp, high confidence
Bug Hunt Bench22.8%estimated ± 7.9 pp, low confidence
Chartography (tools)81.7%estimated ± 4.1 pp, low confidence
CursorBench 4.034.1%measured
CWE-bench v157.4%estimated ± 5.1 pp, medium confidence
DeepSWE65.9%estimated ± 4.1 pp, low confidence
DRACO86.9%estimated ± 0.7 pp, low confidence
FrontierCode 2.027.1%estimated ± 4.7 pp, medium confidence
FrontierPhysics1.7%estimated ± 4.6 pp, high confidence
HLE w/o tools43.2%measured
OSWorld 2.015.9%estimated ± 3.6 pp, high confidence
OSWorld 2.157.0%measured
OSWorld 2.1 Offline54.4%estimated ± 1.9 pp, medium confidence
PostTrainBench v1.135.6%estimated ± 1.7 pp, low confidence
ProgramBench78.3%estimated ± 6.0 pp, low confidence
SpireClimb5.8%estimated ± 7.2 pp, medium confidence
StealthMark41.0%estimated ± 3.5 pp, high confidence
SWE Multilingual78.3%measured
SWE-bench Multimodal28.1%measured
Terminal-Bench 4.019.7%estimated ± 4.5 pp, high confidence
Terminal-Bench-Science 0.10.0%estimated ± 10.7 pp, low confidence
Toolathlon-Verified70.9%estimated ± 2.3 pp, medium confidence
VulcanBench Frontier v482.0%estimated ± 1.8 pp, medium confidence
AA AutomationBench42.0%estimated ± 7.9 pp, low confidence
GDP.pdf19.0%estimated ± 4.2 pp, high confidence
AA Terminal-Bench 4.011.7%estimated ± 6.2 pp, medium confidence
ARC-AGI-35.0%estimated ± 7.2 pp, medium confidence
ARC-AGI-3 Provider Adapter2.7%estimated ± 7.3 pp, medium confidence
cursorBench4135.7%estimated ± 2.0 pp, medium confidence
CWE-bench57.5%estimated ± 4.6 pp, high confidence
MKA34.9%estimated ± 4.6 pp, high confidence
SkillsBench official leaderboard (BenchFlow)55.1%estimated ± 6.1 pp, medium confidence
BinaryAudit35.8%estimated ± 12.1 pp, low confidence
CompileBench100.0%estimated ± 10.0 pp, low confidence
OTelBench22.5%estimated ± 6.6 pp, medium confidence
Vibe Code Bench71.8%estimated ± 5.6 pp, low confidence
FACTS Multimodal42.5%estimated ± 4.0 pp, high confidence
Hermes Bench71.2%estimated ± 1.8 pp, high confidence
Hermes Index37.2%estimated ± 3.1 pp, high confidence
ParseBench-266.8%estimated ± 3.5 pp, high confidence
CharXiv88.3%measured
CharXiv w/o tools77.0%measured
ExploitBench37.6%estimated ± 1.9 pp, medium confidence
ExploitGym12.2%estimated ± 2.7 pp, high confidence
AI2D_TEST93.5%estimated ± 0.9 pp, low confidence
C-Eval97.0%estimated ± 1.0 pp, low confidence
Claw-Eval69.8%estimated ± 6.2 pp, low confidence
CountBench96.5%estimated ± 2.9 pp, medium confidence
DynaMath87.7%estimated ± 0.8 pp, low confidence
ERQA58.0%estimated ± 2.9 pp, medium confidence
HMMT Feb 202694.5%estimated ± 1.5 pp, medium confidence
HMMT Nov 202595.3%estimated ± 1.5 pp, low confidence
INCLUDE86.0%estimated ± 2.0 pp, medium confidence
MathVision87.9%estimated ± 3.1 pp, high confidence
MCP Atlas77.7%estimated ± 5.8 pp, low confidence
MLVU (M-Avg)88.4%estimated ± 1.5 pp, high confidence
MMAnswerBench90.5%estimated ± 2.4 pp, low confidence
MMLU-ProX86.0%estimated ± 1.3 pp, high confidence
MMLU-Redux96.0%estimated ± 0.8 pp, medium confidence
MMMU-Pro78.3%estimated ± 1.5 pp, high confidence
NL2Repo45.2%estimated ± 4.7 pp, high confidence
PostTrainBench v123.9%estimated ± 3.6 pp, high confidence
QwenClawBench93.2%estimated ± 6.0 pp, low confidence
RealWorldQA86.7%estimated ± 1.5 pp, high confidence
ScreenSpot Pro83.2%estimated ± 6.8 pp, medium confidence
SuperGPQA96.2%estimated ± 7.2 pp, low confidence
Terminal-Bench 2.075.8%estimated ± 5.8 pp, medium confidence
Toolathlon54.9%estimated ± 1.7 pp, medium confidence
V*100.0%estimated ± 7.5 pp, medium confidence
Video-MME (w/o subtitle)84.7%estimated ± 1.7 pp, medium confidence
Video-MME (with subtitle)90.1%estimated ± 2.9 pp, high confidence
VideoMMMU86.8%estimated ± 1.5 pp, medium confidence
We-Math84.6%estimated ± 7.9 pp, low confidence
AA LiveCodeBench91.4%estimated ± 2.6 pp, medium confidence
HealthBench Hard31.8%estimated ± 8.6 pp, low confidence
LiveCodeBench Pro86.2%estimated ± 6.0 pp, low confidence
MedXpertQA (MM)78.4%estimated ± 6.4 pp, low confidence
MedXpertQA (Text)60.7%estimated ± 8.9 pp, low confidence
React Native Evals82.0%estimated ± 2.7 pp, medium confidence
ResearchClawBench20.7%estimated ± 1.9 pp, medium confidence
SWE-Rebench61.8%estimated ± 3.6 pp, low confidence
PostTrainBench21.4%estimated ± 3.7 pp, high confidence
DAYJOB1.7%measured
DAYJOB Healthcare0.0%measured
Finance Agent v255.6%estimated ± 2.4 pp, medium confidence
BioMysteryBench (human-solvable)88.4%estimated ± 0.6 pp, medium confidence
FrontierCode 1.1 Extended55.1%estimated ± 1.6 pp, medium confidence
HealthBench (raw)49.9%estimated ± 6.0 pp, medium confidence
HealthBench (length-adjusted)54.1%estimated ± 3.4 pp, medium confidence
HealthBench Professional52.4%estimated ± 4.2 pp, medium confidence
HealthBench Professional (raw)60.4%estimated ± 4.5 pp, high confidence
HLE-Verified31.0%measured
LABBench280.1%measured
ScarfBench45.6%estimated ± 9.1 pp, low confidence
BlenderWorld (text-only)72.1%estimated ± 2.8 pp, medium confidence
Cookie-Bench81.5%estimated ± 1.3 pp, medium confidence
STEMBench2.5%estimated ± 0.2 pp, medium confidence
SimpleQA45.7%estimated ± 4.3 pp, low confidence
LiveCodeBench82.3%estimated ± 14.6 pp, low confidence
Agents' Last Exam30.9%estimated ± 9.7 pp, medium confidence
BabyVision w/ Python85.0%estimated ± 3.7 pp, low confidence
OpenHarmony Bench v1.058.9%estimated ± 2.4 pp, high confidence
SEC-Bench Pro46.4%estimated ± 5.8 pp, medium confidence
BabyVision42.3%estimated ± 7.4 pp, low confidence
MMVU79.9%estimated ± 3.6 pp, medium confidence
PerceptionBench55.3%estimated ± 3.8 pp, low confidence
ZeroBench14.7%estimated ± 7.1 pp, low confidence
MRCRv285.9%estimated ± 11.3 pp, low confidence
LVBench62.8%estimated ± 8.2 pp, low confidence
MMLongBench-Doc61.5%estimated ± 2.6 pp, low confidence
BioBasicsBench34.0%estimated ± 2.8 pp, medium confidence
LifeVal75.9%estimated ± 1.6 pp, medium confidence
MMMU-Pro w/ Python80.3%estimated ± 2.0 pp, high confidence
ExploitBench Internal Port0.9%estimated ± 1.4 pp, medium confidence
MathVision w/ Python94.8%estimated ± 1.0 pp, low confidence
PinchBench90.4%estimated ± 3.6 pp, low confidence