benchgap
Google · model

Gemini 3.6 Flash benchmark scores

As of 2026-11-22, Gemini 3.6 Flash (Google) has measured scores on 34 benchmarks and estimated scores on 141 more.

BenchmarkScoreSource
AA-GPQA Diamond92.8%measured
AA-HLE40.8%measured
AA-SciCode53.4%measured
AIME2695.9%estimated ± 1.2 pp, medium confidence
Artificial Analysis Intelligence Index34.0%measured
CritPt10.6%measured
GDPval-AA39.2%measured
GPQA Diamond91.4%estimated ± 1.4 pp, high confidence
HLE46.0%estimated ± 6.1 pp, medium confidence
AA-LCR80.0%measured
LiveCodeBench v690.1%estimated ± 2.2 pp, low confidence
AA-Omniscience Accuracy50.0%measured
SciCode52.9%estimated ± 2.9 pp, low confidence
Terminal-Bench 2.185.5%estimated ± 6.7 pp, medium confidence
BrowseComp83.8%estimated ± 4.7 pp, high confidence
HLE w/ tools52.7%estimated ± 5.3 pp, medium confidence
LongBench v265.7%estimated ± 5.5 pp, medium confidence
AA Coding Index69.2%measured
AA-MMMU-Pro83.2%measured
APEX-Agents31.1%estimated ± 9.4 pp, low confidence
APEX-Agents-AA45.7%estimated ± 7.8 pp, low confidence
BioMysteryBench (human-difficult)44.3%estimated ± 6.8 pp, low confidence
DeepSearchQA89.2%estimated ± 4.5 pp, high confidence
SWE-bench Verified81.7%estimated ± 4.4 pp, high confidence
BFCL v475.7%estimated ± 2.7 pp, medium confidence
CyberGym68.7%estimated ± 5.3 pp, medium confidence
JobBench39.9%estimated ± 8.7 pp, medium confidence
SWE-bench Pro57.0%estimated ± 4.5 pp, high confidence
WideResearch81.4%estimated ± 2.1 pp, medium confidence
AA Agentic Index30.2%measured
MMLU-Pro87.3%estimated ± 4.8 pp, high confidence
FailureSensorIQ68.0%estimated ± 1.7 pp, high confidence
GPQA91.2%estimated ± 1.6 pp, high confidence
MMLU90.5%estimated ± 0.9 pp, medium confidence
Vals GPQA Diamond93.4%measured
Vals LiveCodeBench88.1%measured
Vals MMLU-Pro89.3%measured
MMMLU89.6%estimated ± 0.7 pp, medium confidence
MMMU88.3%estimated ± 2.7 pp, medium confidence
Chess Puzzles40.0%measured
AA MMLU-Pro88.8%estimated ± 0.7 pp, medium confidence
Gert Labs59.5%measured
AA-AnalystAgent44.8%estimated ± 5.9 pp, low confidence
AA Harvey LAB v1.087.7%estimated ± 1.2 pp, medium confidence
AA Tau3 Banking33.3%estimated ± 3.7 pp, high confidence
AA Terminal-Bench 2.183.9%estimated ± 2.5 pp, high confidence
ApprenticeBench25.2%estimated ± 6.3 pp, low confidence
ARC-AGI-191.2%measured
ARC-AGI-260.4%measured
BabaIsBench88.0%measured
CursorBench 3.152.4%estimated ± 3.4 pp, medium confidence
CursorBench 3.253.5%measured
EBR-bench9.1%estimated ± 9.3 pp, medium confidence
Terminal-Bench 3.017.7%estimated ± 6.9 pp, medium confidence
FrontierCode 1.1 Main36.5%estimated ± 1.6 pp, low confidence
FrontierMath v2 (Tier 4)22.0%measured
FrontierMath v2 (Tiers 1-3)58.9%measured
FrontierSWE v210.1%estimated ± 9.5 pp, medium confidence
Furniture Assembly23.3%measured
MaxxBench49.4%estimated ± 6.6 pp, low confidence
MazeBench0.0%measured
MirrorCode23.3%measured
Mystery Game Puzzles30.0%measured
NYT Connections89.0%measured
OfficeQA Pro67.6%estimated ± 6.8 pp, medium confidence
OSWorld-Verified83.0%measured
ParseBench-169.1%measured
Terminal-Bench 4.122.8%estimated ± 14.9 pp, low confidence
Terminal-Bench Hard56.4%estimated ± 3.9 pp, high confidence
Vals SWE-bench79.6%measured
Terminal-Bench 2.1 (Vals AI)73.8%measured
VulcanBench v375.4%estimated ± 5.1 pp, low confidence
CWE-Bench-AA56.0%estimated ± 8.1 pp, medium confidence
AA Cyber Index19.3%estimated ± 9.2 pp, low confidence
DeepsecBench-AA22.0%estimated ± 8.2 pp, medium confidence
HealthBench Professional (length-adjusted, Artificial Analysis)56.2%estimated ± 4.3 pp, low confidence
HealthBench Professional (raw, Artificial Analysis)56.1%estimated ± 0.8 pp, low confidence
AA ITBench41.0%estimated ± 3.2 pp, high confidence
Bug Hunt Bench18.2%estimated ± 9.0 pp, low confidence
Chartography (tools)85.3%estimated ± 4.1 pp, low confidence
CursorBench 4.031.9%estimated ± 2.0 pp, medium confidence
CWE-bench v157.4%estimated ± 5.1 pp, medium confidence
DeepSWE49.0%measured
DRACO86.9%estimated ± 0.7 pp, low confidence
FrontierCode 2.016.2%estimated ± 5.4 pp, low confidence
FrontierPhysics0.0%estimated ± 4.4 pp, high confidence
HLE w/o tools48.2%estimated ± 4.9 pp, high confidence
OSWorld 2.017.8%estimated ± 3.6 pp, high confidence
OSWorld 2.172.9%estimated ± 3.6 pp, high confidence
OSWorld 2.1 Offline26.6%estimated ± 1.9 pp, low confidence
PostTrainBench v1.123.8%estimated ± 3.7 pp, high confidence
ProgramBench62.2%estimated ± 6.0 pp, low confidence
SpireClimb5.8%estimated ± 7.2 pp, medium confidence
StealthMark45.2%estimated ± 6.5 pp, medium confidence
SWE-bench Multimodal56.6%estimated ± 5.6 pp, low confidence
Toolathlon-Verified60.0%estimated ± 2.3 pp, medium confidence
VulcanBench Frontier v470.7%estimated ± 2.3 pp, low confidence
AA AutomationBench59.4%estimated ± 9.9 pp, low confidence
GDP.pdf22.4%estimated ± 4.2 pp, high confidence
AA Terminal-Bench 4.035.3%estimated ± 13.7 pp, low confidence
ARC-AGI-35.5%estimated ± 8.6 pp, low confidence
ARC-AGI-3 Provider Adapter2.7%estimated ± 7.3 pp, medium confidence
cursorBench4144.8%estimated ± 3.7 pp, medium confidence
CWE-bench52.7%estimated ± 6.2 pp, medium confidence
MKA39.7%estimated ± 4.6 pp, high confidence
SkillsBench official leaderboard (BenchFlow)55.9%estimated ± 6.1 pp, medium confidence
BinaryAudit35.8%estimated ± 12.1 pp, low confidence
CompileBench100.0%estimated ± 10.0 pp, low confidence
OTelBench32.1%estimated ± 4.1 pp, low confidence
Vibe Code Bench51.8%estimated ± 10.0 pp, low confidence
FACTS Multimodal53.7%measured
Hermes Bench65.5%estimated ± 2.9 pp, high confidence
Hermes Index37.2%estimated ± 3.1 pp, high confidence
ParseBench-275.8%measured
CharXiv87.1%estimated ± 5.6 pp, low confidence
CharXiv w/o tools84.3%estimated ± 1.8 pp, medium confidence
ExploitBench32.4%estimated ± 6.5 pp, low confidence
ExploitGym10.2%estimated ± 2.7 pp, high confidence
AI2D_TEST94.2%estimated ± 2.0 pp, low confidence
C-Eval95.7%estimated ± 1.0 pp, low confidence
Claw-Eval70.8%estimated ± 6.2 pp, low confidence
CountBench100.0%estimated ± 2.9 pp, low confidence
DynaMath85.5%estimated ± 3.0 pp, low confidence
ERQA68.8%estimated ± 3.1 pp, medium confidence
HMMT Feb 202691.8%estimated ± 1.5 pp, medium confidence
HMMT Nov 202595.0%estimated ± 1.5 pp, medium confidence
MathVision93.2%estimated ± 3.1 pp, medium confidence
MCP Atlas78.7%estimated ± 7.8 pp, low confidence
MMAnswerBench88.8%estimated ± 2.4 pp, medium confidence
MMLU-Redux97.2%estimated ± 0.8 pp, low confidence
MMMU-Pro83.2%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
RealWorldQA88.1%estimated ± 2.9 pp, low confidence
ScreenSpot Pro88.7%estimated ± 10.1 pp, low confidence
SuperGPQA99.1%estimated ± 9.3 pp, low confidence
Terminal-Bench 2.075.4%estimated ± 5.8 pp, medium confidence
Toolathlon55.8%estimated ± 6.3 pp, medium confidence
V*100.0%estimated ± 7.5 pp, low confidence
Video-MME (w/o subtitle)89.1%estimated ± 1.7 pp, low confidence
Video-MME (with subtitle)90.2%estimated ± 3.7 pp, low confidence
AA LiveCodeBench91.4%estimated ± 2.6 pp, medium confidence
HealthBench Hard29.3%estimated ± 8.6 pp, low confidence
MedXpertQA (MM)77.6%estimated ± 6.5 pp, low confidence
React Native Evals81.5%estimated ± 3.2 pp, high confidence
ResearchClawBench18.8%estimated ± 1.9 pp, high confidence
PostTrainBench21.4%estimated ± 3.7 pp, high confidence
DAYJOB1.4%estimated ± 2.7 pp, medium confidence
DAYJOB Healthcare0.0%estimated ± 4.0 pp, medium confidence
Finance Agent v257.1%estimated ± 2.4 pp, medium confidence
BioMysteryBench (human-solvable)83.8%estimated ± 0.6 pp, low confidence
FrontierCode 1.1 Extended55.1%estimated ± 1.6 pp, low confidence
HealthBench (raw)49.4%estimated ± 6.0 pp, medium confidence
HealthBench (length-adjusted)55.6%estimated ± 2.5 pp, low confidence
HealthBench Professional57.2%estimated ± 4.2 pp, medium confidence
HealthBench Professional (raw)60.0%estimated ± 4.5 pp, high confidence
HLE-Verified52.7%estimated ± 2.0 pp, medium confidence
LABBench282.7%estimated ± 1.8 pp, low confidence
ScarfBench61.3%estimated ± 3.8 pp, low confidence
BlenderWorld (text-only)70.9%estimated ± 2.8 pp, medium confidence
Cookie-Bench79.2%estimated ± 1.3 pp, medium confidence
STEMBench2.6%estimated ± 0.2 pp, medium confidence
SimpleQA35.0%estimated ± 7.8 pp, low confidence
Agents' Last Exam34.6%estimated ± 9.7 pp, medium confidence
BabyVision w/ Python92.8%estimated ± 4.0 pp, low confidence
OpenHarmony Bench v1.058.9%estimated ± 2.4 pp, high confidence
SEC-Bench Pro84.1%estimated ± 5.6 pp, medium confidence
MRCRv285.8%estimated ± 11.3 pp, low confidence
MMLongBench-Doc61.4%estimated ± 2.6 pp, low confidence
BioBasicsBench34.0%estimated ± 2.8 pp, medium confidence
LifeVal75.2%estimated ± 3.8 pp, medium confidence
MMMU-Pro w/ Python85.6%estimated ± 2.0 pp, high confidence
ExploitBench Internal Port0.4%estimated ± 1.4 pp, medium confidence
SWE-Atlas Codebase QnA44.1%estimated ± 4.8 pp, high confidence