benchgap
Google · model

Gemini 3.1 Pro benchmark scores

As of 2026-11-21, Gemini 3.1 Pro (Google) has measured scores on 60 benchmarks and estimated scores on 135 more.

BenchmarkScoreSource
AA-GPQA Diamond94.1%measured
AA-HLE47.0%measured
AA-SciCode58.7%measured
AIME2697.7%estimated ± 1.2 pp, medium confidence
Artificial Analysis Intelligence Index36.2%estimated ± 5.6 pp, medium confidence
CritPt17.7%measured
GDPval-AA14.6%measured
GPQA Diamond94.3%measured
HLE54.3%estimated ± 2.2 pp, high confidence
IFBench75.5%estimated ± 6.6 pp, medium confidence
AA-LCR82.0%measured
LiveCodeBench v690.3%estimated ± 2.2 pp, low confidence
AA-Omniscience Accuracy58.4%estimated ± 6.5 pp, medium confidence
SciCode53.8%estimated ± 2.9 pp, low confidence
Terminal-Bench 2.173.0%estimated ± 5.6 pp, medium confidence
BrowseComp84.5%estimated ± 4.2 pp, high confidence
HLE w/ tools56.9%estimated ± 4.1 pp, medium confidence
IFEval93.1%estimated ± 2.8 pp, high confidence
LongBench v267.4%estimated ± 5.5 pp, medium confidence
VITA-Bench35.6%estimated ± 10.7 pp, low confidence
τ²-bench results95.6%measured
AA Coding Index68.8%measured
AA-MMMU-Pro82.4%measured
APEX-Agents33.2%estimated ± 4.9 pp, low confidence
APEX-Agents-AA32.0%measured
BioMysteryBench (human-difficult)47.5%estimated ± 6.8 pp, low confidence
DeepSearchQA69.7%measured
SWE-bench Verified82.9%estimated ± 3.2 pp, medium confidence
BFCL v470.1%estimated ± 2.7 pp, high confidence
CyberGym53.5%estimated ± 5.3 pp, low confidence
JobBench39.9%estimated ± 8.2 pp, low confidence
SWE-bench Pro55.5%estimated ± 2.9 pp, medium confidence
WideResearch77.3%estimated ± 5.9 pp, low confidence
AA Agentic Index10.3%measured
MMLU-Pro85.7%estimated ± 2.5 pp, low confidence
AA-IFBench77.1%measured
FailureSensorIQ68.6%estimated ± 1.7 pp, high confidence
GPQA94.3%estimated ± 0.0 pp, high confidence
MMLU96.9%estimated ± 1.2 pp, medium confidence
Vals GPQA Diamond95.5%measured
Vals LiveCodeBench88.5%measured
Vals MMLU-Pro91.0%measured
MMMLU90.6%estimated ± 0.7 pp, low confidence
MMMU88.0%estimated ± 0.8 pp, medium confidence
Chess Puzzles51.0%measured
AA MMLU-Pro89.6%estimated ± 1.1 pp, medium confidence
Gert Labs56.9%measured
AA-AnalystAgent51.6%estimated ± 5.9 pp, low confidence
AA EnterpriseOps-Gym42.2%measured
AA Harvey LAB v1.079.4%estimated ± 1.2 pp, low confidence
AA Tau3 Banking39.2%estimated ± 2.7 pp, medium confidence
AA Terminal-Bench 2.182.9%estimated ± 2.5 pp, high confidence
ApprenticeBench0.0%estimated ± 3.1 pp, low confidence
ARC-AGI-194.1%estimated ± 1.7 pp, high confidence
ARC-AGI-277.1%measured
BabaIsBench92.0%measured
CursorBench 3.165.5%estimated ± 3.4 pp, medium confidence
CursorBench 3.255.8%estimated ± 4.4 pp, high confidence
EBR-bench14.3%measured
Terminal-Bench 3.08.9%estimated ± 3.2 pp, medium confidence
FrontierCode 1.1 Main40.7%estimated ± 2.8 pp, low confidence
FrontierMath v2 (Tier 4)26.8%measured
FrontierMath v2 (Tiers 1-3)59.6%measured
FrontierSWE v29.7%estimated ± 9.5 pp, medium confidence
Furniture Assembly26.7%measured
MaxxBench42.0%estimated ± 4.3 pp, low confidence
MazeBench1.0%measured
MirrorCode8.9%measured
Mystery Game Puzzles34.0%measured
NYT Connections97.4%measured
OfficeQA Pro67.3%estimated ± 6.8 pp, medium confidence
OSWorld-Verified67.9%estimated ± 3.8 pp, medium confidence
ParseBench-167.3%measured
Terminal-Bench 4.12.7%measured
Terminal-Bench Hard53.8%measured
Vals SWE-bench78.8%measured
Terminal-Bench 2.1 (Vals AI)70.8%measured
VulcanBench v386.2%estimated ± 5.6 pp, low confidence
CWE-Bench-AA20.9%estimated ± 10.1 pp, low confidence
AA Cyber Index7.3%estimated ± 9.2 pp, low confidence
DeepsecBench-AA0.1%estimated ± 8.9 pp, low confidence
HealthBench Professional (length-adjusted, Artificial Analysis)58.1%estimated ± 4.4 pp, low confidence
HealthBench Professional (raw, Artificial Analysis)63.9%estimated ± 0.8 pp, low confidence
AA ITBench39.3%estimated ± 2.8 pp, high confidence
Bug Hunt Bench5.0%estimated ± 6.3 pp, low confidence
Chartography (tools)84.8%estimated ± 4.1 pp, low confidence
CursorBench 4.034.1%estimated ± 2.0 pp, medium confidence
CWE-bench v157.4%estimated ± 5.1 pp, medium confidence
DeepSWE54.4%estimated ± 4.3 pp, medium confidence
DRACO86.9%estimated ± 0.7 pp, low confidence
FrontierCode 2.019.1%estimated ± 6.0 pp, low confidence
FrontierPhysics0.0%estimated ± 4.4 pp, high confidence
HLE w/o tools45.4%measured
OSWorld 2.05.9%estimated ± 5.9 pp, low confidence
OSWorld 2.178.6%estimated ± 6.1 pp, medium confidence
OSWorld 2.1 Offline68.5%estimated ± 13.8 pp, low confidence
PostTrainBench v1.122.0%measured
ProgramBench0.0%estimated ± 4.8 pp, low confidence
SpireClimb16.8%estimated ± 7.2 pp, medium confidence
StealthMark40.1%estimated ± 2.8 pp, medium confidence
SWE-bench Multimodal55.7%estimated ± 5.6 pp, low confidence
Terminal-Bench 4.011.5%estimated ± 4.5 pp, medium confidence
Toolathlon-Verified31.1%estimated ± 2.3 pp, low confidence
VulcanBench Frontier v468.9%estimated ± 2.8 pp, low confidence
AA AutomationBench55.0%estimated ± 12.8 pp, low confidence
GDP.pdf17.0%estimated ± 4.6 pp, medium confidence
AA Terminal-Bench 4.02.3%estimated ± 6.2 pp, low confidence
ARC-AGI-30.4%measured
ARC-AGI-3 Provider Adapter2.2%estimated ± 4.4 pp, high confidence
cursorBench4145.0%estimated ± 3.7 pp, high confidence
CWE-bench54.7%estimated ± 7.9 pp, low confidence
MKA42.4%estimated ± 5.7 pp, low confidence
SkillsBench official leaderboard (BenchFlow)60.8%measured
BinaryAudit49.1%measured
CompileBench100.0%estimated ± 10.0 pp, low confidence
OTelBench20.1%estimated ± 5.6 pp, low confidence
Vibe Code Bench32.0%measured
FACTS Multimodal50.4%measured
Hermes Bench43.2%estimated ± 2.9 pp, medium confidence
Hermes Index37.2%estimated ± 3.1 pp, high confidence
ParseBench-273.3%estimated ± 3.5 pp, high confidence
CharXiv80.2%measured
CharXiv w/o tools88.9%estimated ± 1.2 pp, high confidence
ExploitBench31.5%estimated ± 1.9 pp, low confidence
ExploitGym6.6%estimated ± 2.7 pp, medium confidence
AI2D_TEST93.4%estimated ± 0.9 pp, medium confidence
C-Eval94.9%estimated ± 1.2 pp, low confidence
Claw-Eval57.8%measured
CountBench97.3%estimated ± 2.2 pp, medium confidence
DynaMath85.7%estimated ± 0.8 pp, medium confidence
ERQA69.4%measured
HMMT Feb 202693.6%estimated ± 1.5 pp, medium confidence
HMMT Nov 202595.0%estimated ± 1.5 pp, low confidence
MathVision100.0%estimated ± 2.2 pp, low confidence
MCP Atlas64.6%estimated ± 7.8 pp, low confidence
MCP-Tasks69.3%estimated ± 6.4 pp, low confidence
MLVU (M-Avg)85.7%estimated ± 1.5 pp, high confidence
MMAnswerBench89.0%estimated ± 2.4 pp, low confidence
MMLU-Redux93.2%estimated ± 2.8 pp, medium confidence
MMMU-Pro83.9%measured
NL2Repo44.7%estimated ± 4.7 pp, high confidence
OmniDocBench 1.590.0%estimated ± 2.0 pp, medium confidence
PostTrainBench v121.6%measured
QwenClawBench55.9%estimated ± 6.0 pp, low confidence
RealWorldQA87.7%estimated ± 1.4 pp, medium confidence
ScreenSpot Pro84.4%measured
SimpleVQA72.4%measured
SuperGPQA83.0%estimated ± 13.9 pp, low confidence
Terminal-Bench 2.070.7%estimated ± 5.4 pp, medium confidence
Toolathlon27.0%estimated ± 6.3 pp, low confidence
V*96.1%estimated ± 6.5 pp, low confidence
Video-MME (w/o subtitle)87.4%estimated ± 1.3 pp, medium confidence
Video-MME (with subtitle)86.6%estimated ± 2.9 pp, high confidence
VideoMMMU85.3%estimated ± 1.5 pp, medium confidence
We-Math87.2%estimated ± 2.1 pp, medium confidence
AA LiveCodeBench90.9%estimated ± 2.6 pp, medium confidence
HealthBench Hard20.6%measured
LiveCodeBench Pro82.9%measured
MedXpertQA (MM)81.3%measured
MedXpertQA (Text)71.5%measured
MMLU-Pro (Arcee)88.3%estimated ± 3.4 pp, low confidence
React Native Evals78.9%measured
ResearchClawBench13.3%measured
SWE-Rebench58.9%estimated ± 6.5 pp, low confidence
PostTrainBench18.1%measured
DAYJOB0.0%measured
DAYJOB Healthcare0.0%measured
Finance Agent v256.5%estimated ± 3.8 pp, medium confidence
BioMysteryBench (human-solvable)88.7%estimated ± 0.6 pp, medium confidence
FrontierCode 1.1 Extended56.0%estimated ± 3.4 pp, low confidence
HealthBench (raw)54.9%estimated ± 6.0 pp, medium confidence
HealthBench (length-adjusted)53.8%estimated ± 0.9 pp, medium confidence
HealthBench Professional56.8%estimated ± 3.6 pp, high confidence
HealthBench Professional (raw)65.2%estimated ± 4.5 pp, high confidence
HLE-Verified49.0%estimated ± 1.3 pp, low confidence
LABBench283.6%estimated ± 1.8 pp, low confidence
ScarfBench37.6%measured
BlenderWorld (text-only)71.8%estimated ± 2.8 pp, medium confidence
Cookie-Bench81.0%estimated ± 1.3 pp, medium confidence
SimpleQA35.7%estimated ± 7.8 pp, low confidence
Agents' Last Exam36.5%estimated ± 11.0 pp, low confidence
BabyVision w/ Python95.2%estimated ± 3.8 pp, low confidence
OpenHarmony Bench v1.058.9%estimated ± 2.4 pp, high confidence
SEC-Bench Pro81.7%estimated ± 5.6 pp, medium confidence
BabyVision73.6%estimated ± 9.1 pp, low confidence
MMVU80.7%estimated ± 2.7 pp, low confidence
PerceptionBench68.3%estimated ± 2.7 pp, medium confidence
ZeroBench29.0%measured
MRCRv285.9%estimated ± 11.3 pp, low confidence
MMLongBench-Doc61.5%estimated ± 2.6 pp, low confidence
OCRBench V272.6%estimated ± 3.3 pp, medium confidence
BioBasicsBench36.9%estimated ± 2.8 pp, medium confidence
LifeVal75.2%estimated ± 3.8 pp, medium confidence
MMMU-Pro w/ Python84.9%estimated ± 2.0 pp, high confidence
ExploitBench Internal Port2.2%estimated ± 1.4 pp, medium confidence