benchgap
xAI · model

Grok 4.6 benchmark scores

As of 2026-11-22, Grok 4.6 (xAI) has measured scores on 47 benchmarks and estimated scores on 109 more.

BenchmarkScoreSource
AA-GPQA Diamond94.9%measured
AA-HLE42.9%measured
AA-SciCode56.5%measured
AIME2699.2%estimated ± 1.2 pp, low confidence
Artificial Analysis Intelligence Index44.3%measured
CritPt17.1%measured
GDPval-AA55.9%measured
GPQA Diamond90.1%estimated ± 1.4 pp, high confidence
HLE54.8%estimated ± 2.2 pp, high confidence
AA-LCR80.3%measured
LiveCodeBench v690.1%estimated ± 2.2 pp, low confidence
AA-Omniscience Accuracy48.2%measured
SciCode53.1%estimated ± 2.9 pp, low confidence
Terminal-Bench 2.185.4%estimated ± 5.6 pp, medium confidence
BrowseComp84.1%estimated ± 2.9 pp, high confidence
HLE w/ tools60.8%estimated ± 5.3 pp, medium confidence
LongBench v265.9%estimated ± 5.5 pp, medium confidence
τ²-bench results91.9%estimated ± 6.5 pp, low confidence
AA Coding Index76.8%measured
APEX-Agents57.5%measured
APEX-Agents-AA36.9%estimated ± 3.5 pp, low confidence
BioMysteryBench (human-difficult)47.3%estimated ± 6.8 pp, low confidence
DeepSearchQA93.8%estimated ± 4.5 pp, high confidence
SWE-bench Verified89.9%estimated ± 4.4 pp, high confidence
AutomationBench48.5%estimated ± 7.1 pp, low confidence
BFCL v477.1%estimated ± 2.7 pp, medium confidence
CyberGym85.4%estimated ± 3.1 pp, low confidence
JobBench60.7%estimated ± 8.7 pp, medium confidence
SWE-bench Pro67.0%estimated ± 4.2 pp, high confidence
WideResearch85.3%estimated ± 3.3 pp, medium confidence
AA Agentic Index53.4%measured
MMLU-Pro87.3%estimated ± 4.8 pp, medium confidence
FailureSensorIQ67.3%estimated ± 1.7 pp, high confidence
GPQA93.2%estimated ± 0.9 pp, high confidence
MMLU90.5%estimated ± 0.9 pp, medium confidence
Vals GPQA Diamond94.7%measured
Vals LiveCodeBench88.2%measured
Vals MMLU-Pro89.4%measured
MMMLU90.5%estimated ± 0.7 pp, low confidence
Chess Puzzles33.0%measured
AA MMLU-Pro88.6%estimated ± 0.7 pp, medium confidence
Gert Labs69.2%measured
AA-AnalystAgent51.6%estimated ± 5.9 pp, low confidence
AA EnterpriseOps-Gym48.3%measured
AA Harvey LAB v1.093.6%estimated ± 1.2 pp, medium confidence
AA Tau3 Banking50.7%measured
AA Terminal-Bench 2.188.4%measured
ApprenticeBench13.0%measured
ARC-AGI-187.0%measured
ARC-AGI-267.1%measured
BabaIsBench96.0%measured
CursorBench 3.160.0%estimated ± 3.4 pp, medium confidence
CursorBench 3.270.8%measured
EBR-bench30.5%measured
Terminal-Bench 3.026.5%measured
FrontierCode 1.1 Main54.7%estimated ± 1.6 pp, low confidence
FrontierMath v2 (Tier 4)31.7%measured
FrontierMath v2 (Tiers 1-3)66.0%measured
FrontierSWE v225.3%measured
Furniture Assembly40.0%measured
MaxxBench58.7%estimated ± 4.3 pp, low confidence
MazeBench1.0%measured
MirrorCode41.1%measured
Mystery Game Puzzles34.0%measured
NYT Connections80.0%measured
OSWorld-Verified77.3%estimated ± 2.3 pp, medium confidence
Terminal-Bench 4.119.7%measured
Terminal-Bench Hard57.9%estimated ± 2.8 pp, medium confidence
Vals SWE-bench95.6%measured
Terminal-Bench 2.1 (Vals AI)78.3%measured
VulcanBench v387.0%measured
CWE-Bench-AA59.4%estimated ± 8.1 pp, medium confidence
AA Cyber Index30.7%estimated ± 9.2 pp, low confidence
DeepsecBench-AA27.6%estimated ± 8.2 pp, medium confidence
HealthBench Professional (length-adjusted, Artificial Analysis)59.3%estimated ± 4.3 pp, medium confidence
HealthBench Professional (raw, Artificial Analysis)63.5%estimated ± 0.8 pp, low confidence
AA ITBench48.8%estimated ± 3.2 pp, high confidence
Bug Hunt Bench27.0%measured
CursorBench 4.041.4%measured
CWE-bench v157.0%measured
DeepSWE65.9%measured
DRACO86.9%estimated ± 0.7 pp, low confidence
FrontierCode 2.027.1%estimated ± 4.7 pp, medium confidence
FrontierPhysics2.4%measured
HLE w/o tools48.6%estimated ± 4.9 pp, high confidence
OSWorld 2.176.9%estimated ± 3.6 pp, high confidence
OSWorld 2.1 Offline72.7%estimated ± 1.9 pp, medium confidence
PostTrainBench v1.135.6%estimated ± 1.7 pp, low confidence
ProgramBench89.4%estimated ± 6.0 pp, low confidence
SpireClimb5.8%estimated ± 7.2 pp, medium confidence
StealthMark49.1%estimated ± 6.0 pp, medium confidence
Terminal-Bench 4.026.3%estimated ± 4.5 pp, high confidence
Terminal-Bench-Science 0.10.0%estimated ± 10.7 pp, low confidence
Toolathlon-Verified75.9%estimated ± 2.3 pp, medium confidence
VulcanBench Frontier v488.5%estimated ± 1.8 pp, high confidence
AA AutomationBench64.7%estimated ± 9.9 pp, low confidence
GDP.pdf24.8%estimated ± 4.2 pp, high confidence
AA Terminal-Bench 4.019.3%estimated ± 6.2 pp, medium confidence
ARC-AGI-32.1%measured
ARC-AGI-3 Provider Adapter10.4%estimated ± 4.1 pp, high confidence
cursorBench4144.3%measured
CWE-bench55.7%estimated ± 5.8 pp, low confidence
MKA44.9%estimated ± 4.6 pp, high confidence
SkillsBench official leaderboard (BenchFlow)54.0%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 Bench78.6%estimated ± 10.0 pp, low confidence
FACTS Multimodal39.3%measured
Hermes Bench73.0%estimated ± 2.4 pp, low confidence
Hermes Index37.2%estimated ± 3.1 pp, high confidence
ExploitBench32.4%estimated ± 6.5 pp, medium confidence
ExploitGym15.8%estimated ± 2.7 pp, high confidence
C-Eval98.8%estimated ± 1.0 pp, low confidence
HMMT Feb 202695.4%estimated ± 1.5 pp, medium confidence
HMMT Nov 202595.3%estimated ± 1.5 pp, low confidence
MCP Atlas76.5%estimated ± 5.8 pp, low confidence
MMAnswerBench90.6%estimated ± 2.4 pp, low confidence
MMLU-Redux97.0%estimated ± 0.8 pp, low confidence
NL2Repo59.6%estimated ± 4.7 pp, high confidence
PostTrainBench v143.7%estimated ± 3.6 pp, medium confidence
SuperGPQA96.2%estimated ± 9.3 pp, low confidence
Terminal-Bench 2.078.1%estimated ± 5.8 pp, low confidence
Toolathlon54.3%estimated ± 1.7 pp, medium confidence
AA LiveCodeBench100.0%estimated ± 2.6 pp, medium confidence
HealthBench Hard31.9%estimated ± 8.6 pp, low confidence
React Native Evals83.4%estimated ± 3.2 pp, medium confidence
ResearchClawBench21.4%estimated ± 1.9 pp, medium confidence
PostTrainBench33.1%estimated ± 2.4 pp, high confidence
DAYJOB1.4%estimated ± 2.7 pp, high confidence
DAYJOB Healthcare0.4%estimated ± 4.0 pp, high confidence
Finance Agent v259.8%estimated ± 2.4 pp, medium confidence
BioMysteryBench (human-solvable)88.5%estimated ± 0.6 pp, medium confidence
FrontierCode 1.1 Extended61.3%measured
HealthBench (raw)51.3%estimated ± 6.0 pp, medium confidence
HealthBench (length-adjusted)53.8%estimated ± 0.9 pp, medium confidence
HealthBench Professional57.0%estimated ± 3.6 pp, high confidence
HealthBench Professional (raw)61.8%estimated ± 4.5 pp, high confidence
HLE-Verified54.5%estimated ± 2.0 pp, medium confidence
LABBench282.8%estimated ± 1.8 pp, low confidence
ScarfBench64.9%estimated ± 3.8 pp, medium confidence
BlenderWorld (text-only)71.8%estimated ± 2.8 pp, medium confidence
Cookie-Bench81.0%estimated ± 1.3 pp, medium confidence
STEMBench2.6%estimated ± 0.2 pp, medium confidence
SimpleQA36.1%estimated ± 7.8 pp, low confidence
Agents' Last Exam40.6%estimated ± 9.7 pp, medium confidence
OpenHarmony Bench v1.058.9%estimated ± 2.4 pp, high confidence
OpenHarmony Bench v1.263.8%estimated ± 11.0 pp, low confidence
SEC-Bench Pro38.3%estimated ± 5.6 pp, medium confidence
MRCRv285.8%estimated ± 11.3 pp, low confidence
MMLongBench-Doc61.4%estimated ± 2.6 pp, low confidence
SWE-Marathon47.4%estimated ± 2.1 pp, medium confidence
BioBasicsBench34.1%estimated ± 2.8 pp, medium confidence
LifeVal75.2%estimated ± 3.8 pp, medium confidence
ExploitBench Internal Port0.2%estimated ± 1.4 pp, low confidence
SWE-Atlas Codebase QnA59.3%estimated ± 4.8 pp, high confidence