benchgap
Google · model

Gemini 2.5 Pro benchmark scores

As of 2026-11-22, Gemini 2.5 Pro (Google) has measured scores on 27 benchmarks and estimated scores on 123 more.

BenchmarkScoreSource
AA-GPQA Diamond84.4%measured
AA-HLE22.5%measured
AA-SciCode46.3%measured
AIME2695.9%estimated ± 1.2 pp, low confidence
Artificial Analysis Intelligence Index16.1%measured
CritPt2.6%measured
GDPval-AA0.0%measured
GPQA Diamond83.1%estimated ± 1.2 pp, high confidence
HLE18.8%measured
IFBench55.6%estimated ± 6.6 pp, medium confidence
AA-LCR69.0%measured
LiveCodeBench v679.8%estimated ± 4.7 pp, high confidence
AA-Omniscience Accuracy39.1%measured
SciCode46.3%estimated ± 3.1 pp, high confidence
BrowseComp70.7%estimated ± 8.8 pp, medium confidence
HLE w/ tools33.8%estimated ± 5.3 pp, low confidence
IFEval90.9%estimated ± 2.8 pp, high confidence
LongBench v258.0%estimated ± 5.5 pp, medium confidence
τ²-bench results54.1%measured
AA Coding Index33.3%measured
AA-MMMU-Pro74.9%measured
APEX-Agents0.0%estimated ± 9.4 pp, low confidence
APEX-Agents-AA16.3%estimated ± 9.3 pp, medium confidence
BioMysteryBench (human-difficult)29.1%estimated ± 6.8 pp, low confidence
DeepSearchQA48.1%estimated ± 4.5 pp, medium confidence
SWE-bench Verified63.8%measured
BFCL v453.9%estimated ± 2.7 pp, high confidence
CyberGym48.3%estimated ± 5.3 pp, low confidence
JobBench0.0%estimated ± 8.7 pp, low confidence
SWE-bench Pro53.0%estimated ± 2.9 pp, low confidence
WideResearch72.7%estimated ± 3.3 pp, high confidence
AA Agentic Index3.6%measured
MMLU-Pro83.9%estimated ± 4.4 pp, high confidence
AA-IFBench48.7%measured
AIME 202492.0%measured
FailureSensorIQ67.0%measured
GPQA83.0%measured
MMLU90.5%estimated ± 0.9 pp, medium confidence
Vals GPQA Diamond85.1%estimated ± 3.1 pp, medium confidence
Vals LiveCodeBench81.1%estimated ± 5.7 pp, low confidence
Vals MMLU-Pro84.2%estimated ± 1.8 pp, high confidence
MATH-50097.7%estimated ± 1.3 pp, medium confidence
MMMLU87.5%estimated ± 0.7 pp, medium confidence
MMMU83.3%estimated ± 2.7 pp, high confidence
AIME 202589.4%estimated ± 3.5 pp, high confidence
AA MMLU-Pro86.2%measured
Gert Labs32.6%measured
AA Harvey LAB v1.079.4%estimated ± 1.2 pp, low confidence
AA Tau3 Banking18.1%estimated ± 3.7 pp, medium confidence
ApprenticeBench3.4%estimated ± 10.4 pp, low confidence
ARC-AGI-160.7%estimated ± 6.7 pp, medium confidence
ARC-AGI-28.6%estimated ± 10.9 pp, low confidence
CursorBench 3.134.9%estimated ± 3.4 pp, low confidence
CursorBench 3.247.8%estimated ± 5.3 pp, low confidence
FrontierCode 1.1 Main0.0%estimated ± 4.2 pp, low confidence
FrontierMath v2 (Tier 4)0.0%measured
FrontierMath v2 (Tiers 1-3)24.6%measured
FrontierSWE v212.3%estimated ± 14.2 pp, low confidence
MazeBench0.1%estimated ± 2.7 pp, high confidence
Mystery Game Puzzles6.4%estimated ± 4.7 pp, medium confidence
OfficeQA Pro50.3%estimated ± 6.8 pp, low confidence
OSWorld-Verified55.4%estimated ± 8.4 pp, low confidence
ParseBench-159.0%estimated ± 6.2 pp, medium confidence
Vals SWE-bench54.4%measured
CWE-Bench-AA42.0%estimated ± 8.1 pp, medium confidence
AA Cyber Index2.6%estimated ± 9.2 pp, low confidence
DeepsecBench-AA10.1%estimated ± 8.2 pp, low confidence
HealthBench Professional (length-adjusted, Artificial Analysis)50.7%estimated ± 4.3 pp, low confidence
HealthBench Professional (raw, Artificial Analysis)29.0%estimated ± 0.8 pp, low confidence
AA ITBench34.0%estimated ± 4.0 pp, medium confidence
Bug Hunt Bench19.1%estimated ± 10.5 pp, low confidence
Chartography (tools)80.2%estimated ± 4.1 pp, low confidence
CursorBench 4.032.6%estimated ± 4.9 pp, medium confidence
CWE-bench v140.9%estimated ± 5.7 pp, medium confidence
DeepSWE49.3%estimated ± 7.5 pp, low confidence
DRACO0.0%estimated ± 2.0 pp, low confidence
FrontierCode 2.026.1%estimated ± 9.3 pp, low confidence
HLE w/o tools28.7%estimated ± 4.9 pp, high confidence
OSWorld 2.117.5%estimated ± 3.6 pp, medium confidence
OSWorld 2.1 Offline42.7%estimated ± 10.1 pp, low confidence
PostTrainBench v1.127.3%estimated ± 7.4 pp, low confidence
Toolathlon-Verified13.0%estimated ± 2.3 pp, low confidence
VulcanBench Frontier v476.5%estimated ± 6.8 pp, low confidence
AA AutomationBench44.7%estimated ± 9.9 pp, low confidence
GDP.pdf14.4%estimated ± 4.2 pp, medium confidence
ARC-AGI-31.3%estimated ± 9.8 pp, medium confidence
ARC-AGI-3 Provider Adapter2.7%estimated ± 7.3 pp, low confidence
cursorBench4144.8%estimated ± 3.7 pp, medium confidence
CWE-bench35.2%estimated ± 6.2 pp, low confidence
MKA33.0%estimated ± 4.6 pp, medium confidence
SkillsBench official leaderboard (BenchFlow)44.1%estimated ± 9.5 pp, low confidence
BinaryAudit21.1%measured
CompileBench53.3%measured
OTelBench19.6%estimated ± 5.6 pp, medium confidence
Vibe Code Bench0.4%measured
Hermes Bench0.0%estimated ± 2.9 pp, medium confidence
Hermes Index22.0%estimated ± 5.7 pp, low confidence
ParseBench-264.4%estimated ± 6.2 pp, medium confidence
CharXiv80.8%estimated ± 6.3 pp, low confidence
CharXiv w/o tools80.1%estimated ± 2.5 pp, high confidence
ExploitGym0.0%estimated ± 8.3 pp, low confidence
AI2D_TEST91.4%estimated ± 2.0 pp, medium confidence
C-Eval88.8%estimated ± 0.9 pp, low confidence
CountBench94.9%estimated ± 2.9 pp, medium confidence
DynaMath85.2%estimated ± 3.0 pp, medium confidence
ERQA62.9%estimated ± 3.1 pp, high confidence
HMMT Feb 202683.5%estimated ± 1.5 pp, low confidence
HMMT Nov 202590.7%estimated ± 1.5 pp, medium confidence
MathVision85.1%estimated ± 3.1 pp, high confidence
MCP Atlas64.6%estimated ± 7.8 pp, low confidence
MMAnswerBench80.1%estimated ± 2.4 pp, medium confidence
MMLU-Redux95.9%estimated ± 0.8 pp, medium confidence
MMMU-Pro76.1%estimated ± 1.5 pp, high confidence
NL2Repo36.3%estimated ± 4.7 pp, medium confidence
PostTrainBench v119.0%estimated ± 3.1 pp, low confidence
QwenClawBench55.9%estimated ± 6.0 pp, low confidence
RealWorldQA82.9%estimated ± 2.9 pp, medium confidence
ScreenSpot Pro74.4%estimated ± 10.8 pp, low confidence
SuperGPQA59.9%estimated ± 7.2 pp, medium confidence
Toolathlon27.0%estimated ± 6.3 pp, low confidence
V*100.0%estimated ± 7.5 pp, medium confidence
Video-MME (w/o subtitle)83.0%estimated ± 1.7 pp, medium confidence
Video-MME (with subtitle)85.0%estimated ± 3.7 pp, medium confidence
AA LiveCodeBench80.1%measured
HealthBench Hard17.7%estimated ± 8.6 pp, low confidence
MedXpertQA (MM)71.9%estimated ± 6.5 pp, low confidence
MMLU-Pro (Arcee)86.9%estimated ± 4.0 pp, medium confidence
React Native Evals72.4%estimated ± 2.7 pp, medium confidence
ResearchClawBench6.0%estimated ± 1.9 pp, medium confidence
SWE-Rebench57.0%estimated ± 3.6 pp, low confidence
PostTrainBench17.0%estimated ± 8.3 pp, low confidence
Finance Agent v255.1%estimated ± 2.4 pp, low confidence
BioMysteryBench (human-solvable)81.7%estimated ± 0.6 pp, low confidence
HealthBench (raw)30.7%estimated ± 6.0 pp, low confidence
HealthBench (length-adjusted)46.9%estimated ± 3.4 pp, low confidence
HealthBench Professional56.7%estimated ± 4.2 pp, low confidence
HealthBench Professional (raw)42.5%estimated ± 4.5 pp, medium confidence
HLE-Verified33.8%estimated ± 7.1 pp, low confidence
ScarfBench24.4%estimated ± 9.1 pp, low confidence
STEMBench2.5%estimated ± 0.2 pp, medium confidence
SimpleQA44.3%estimated ± 4.3 pp, low confidence
LiveCodeBench68.7%estimated ± 14.6 pp, low confidence
Agents' Last Exam29.7%estimated ± 9.7 pp, low confidence
BabyVision w/ Python87.6%estimated ± 4.0 pp, low confidence
OpenHarmony Bench v1.051.0%estimated ± 2.4 pp, high confidence
MRCRv269.6%estimated ± 11.3 pp, low confidence
MMLongBench-Doc60.9%estimated ± 2.6 pp, low confidence
BioBasicsBench33.9%estimated ± 2.8 pp, low confidence
LifeVal10.9%estimated ± 5.0 pp, low confidence
MMMU-Pro w/ Python78.1%estimated ± 2.0 pp, high confidence