Gemma 7B — Benchmarks

Benchmark scores for Gemma 7B aggregated from public leaderboards, with how it ranks among open models. See hardware requirements for what you need to run it.

Overall rank: #31 of 73 open modelscomposite 51.3/100 across 5 benchmarks in 3 categories · methodology

Knowledge

Benchmark	Score	Open rank	All models
HellaSwag	82.2%	#12 / 42	#23 / 76
MMLU-Pro	33.7%	#92 / 119	#218 / 259
MMLU	66.1%	#34 / 76	#77 / 136

Math

Benchmark	Score	Open rank	All models
GSM8K	46.4%	#32 / 59	#45 / 93

Reasoning

Benchmark	Score	Open rank	All models
BIG-Bench Hard	55.1%	#14 / 37	#21 / 50

Scores aggregated from public benchmark sources (each linked from the benchmark pages). llmrun does not run these benchmarks.