Skip to content

Benchmark

LABBench2

Science benchmarks measure how well models answer science questions. Scores are percentages. Higher is better.

Models measured
3
Scores recorded
5
Sources
2
IQ weight
No weight published

Score distribution

85.1%
Mean
86.2%
Median
2.6%
Std dev
82.1%
10th percentile
88.8%
90th percentile
82.1%
Min
82.1%88.8%
The band covers the middle 80% of scores. The line marks the median.

Leaderboard

#ModelScore
1Gemini 4 Argon· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim88.8%
2Gemini 3.8 Flash· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim86.2%
3Gemini 3.8 Flashmacro-average across 11 subtasks· Reported by Google DeepMind· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.86.2%
4Gemini 3.7 Flash· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim82.1%
5Gemini 3.7 Flashoverall· Reported by Google DeepMind· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.82.1%

Data credits: BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links. · Data by LLM Stats