Gemma 3 270M: benchmark results

Provider: Google. Released 2025-03-12. Access: Open.

Unified ELO 1328 ± 1, rank #1308 of 1392 rated models, from 307 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AI Energy Score (Text Generation)5Energy Score (1-5)79.5
AA Omniscience-29.27Score57.5
EuroEval Bosnian Summarization - LR SUM BS20.18Score (%)34.2
EuroEval Serbian Summarization - LR SUM SR19.58Score (%)33.3
EuroEval Lithuanian Common Sense Reasoning0.7Common Sense Reasoning Average Score (%)28
EuroEval Slovene Common Sense Reasoning1.25Common Sense Reasoning Average Score (%)25.8
EuroEval Faroese NLU - FoSent5.04Sentiment classification Score (%)24.8
AA CritPt0Accuracy (%)24.4
EuroEval Icelandic NLU - ScaLA IS0.02Linguistic acceptability Score (%)24.3
AA Omniscience - Software Engineering (SWE) - Go8.33Accuracy (%)23.5
EuroEval Dutch Simplification - Duidelijke Taal38.54Score (%)22.4
EuroEval Albanian Summarization - LR SUM SQ20.74Score (%)21.6

Interactive version: theaggregate.ai/model?slug=gemma-3-270m · How It Works · Data refreshed daily, snapshot 2026-09-05.