benchgap
Google · model

Gemini 4 Neon benchmark scores

As of 2026-11-19, Gemini 4 Neon (Google) has measured scores on 6 benchmarks and estimated scores on 51 more.

BenchmarkScoreSource
AA-GPQA Diamond91.9%estimated ± 11.8 pp, low confidence
AA-HLE58.9%estimated ± 9.2 pp, medium confidence
AA-SciCode63.1%estimated ± 3.5 pp, medium confidence
Artificial Analysis Intelligence Index61.4%measured
CritPt32.2%estimated ± 5.3 pp, medium confidence
GDPval-AA60.9%estimated ± 6.4 pp, low confidence
GPQA Diamond93.1%estimated ± 2.5 pp, high confidence
HLE65.5%estimated ± 12.0 pp, low confidence
AA-Omniscience Accuracy59.0%estimated ± 7.9 pp, medium confidence
Terminal-Bench 2.189.1%estimated ± 5.4 pp, low confidence
HLE w/ tools66.7%estimated ± 6.8 pp, low confidence
AA Coding Index79.1%estimated ± 2.3 pp, low confidence
BioMysteryBench (human-difficult)49.8%estimated ± 8.3 pp, low confidence
AutomationBench53.4%estimated ± 6.1 pp, low confidence
JobBench72.7%estimated ± 11.3 pp, low confidence
SWE-bench Pro87.5%estimated ± 8.4 pp, medium confidence
MMLU-Pro86.3%estimated ± 5.5 pp, low confidence
GPQA93.1%estimated ± 6.9 pp, medium confidence
Vals GPQA Diamond93.1%estimated ± 5.7 pp, low confidence
Vals MMLU-Pro90.1%estimated ± 4.2 pp, medium confidence
AA-AnalystAgent58.1%estimated ± 6.3 pp, low confidence
ARC-AGI-1100.0%estimated ± 7.9 pp, low confidence
ARC-AGI-290.1%estimated ± 14.6 pp, low confidence
CursorBench 3.283.4%estimated ± 4.1 pp, low confidence
FrontierCode 1.1 Main52.3%estimated ± 3.9 pp, medium confidence
FrontierSWE v262.5%estimated ± 8.3 pp, medium confidence
MazeBench17.0%estimated ± 6.8 pp, low confidence
Terminal-Bench 4.162.6%estimated ± 6.0 pp, medium confidence
Terminal-Bench 2.1 (Vals AI)84.0%estimated ± 6.9 pp, low confidence
CWE-Bench-AA63.3%estimated ± 3.8 pp, medium confidence
DeepsecBench-AA47.6%estimated ± 7.5 pp, low confidence
Harvey LAB-AA v1.19.2%estimated ± 3.5 pp, medium confidence
HealthBench Professional (length-adjusted, Artificial Analysis)64.6%estimated ± 4.3 pp, medium confidence
HealthBench Professional (raw, Artificial Analysis)71.2%estimated ± 1.1 pp, medium confidence
Bug Hunt Bench51.4%estimated ± 8.8 pp, medium confidence
CursorBench 4.058.7%estimated ± 3.3 pp, high confidence
FrontierCode 2.046.2%measured
HLE w/o tools65.0%estimated ± 5.0 pp, high confidence
OSWorld 2.1 Offline83.8%estimated ± 7.4 pp, medium confidence
PostTrainBench v1.147.1%estimated ± 4.8 pp, high confidence
Terminal-Bench 4.067.2%estimated ± 6.1 pp, medium confidence
Toolathlon-Verified78.0%estimated ± 1.8 pp, low confidence
VulcanBench Frontier v492.4%estimated ± 6.0 pp, low confidence
AA AutomationBench79.3%measured
GDP.pdf28.0%measured
AA Terminal-Bench 4.063.6%measured
CWE-bench75.6%estimated ± 5.0 pp, medium confidence
MKA56.5%measured
Hermes Bench100.0%estimated ± 1.8 pp, low confidence
Hermes Index90.8%estimated ± 2.7 pp, low confidence
ExploitGym50.0%estimated ± 7.2 pp, low confidence
PostTrainBench42.8%estimated ± 4.6 pp, high confidence
HealthBench (length-adjusted)60.7%estimated ± 3.7 pp, medium confidence
HealthBench Professional65.9%estimated ± 4.2 pp, medium confidence
ScarfBench69.9%estimated ± 5.0 pp, low confidence
Agents' Last Exam52.3%estimated ± 10.7 pp, low confidence
BioBasicsBench58.4%estimated ± 10.2 pp, low confidence