granite-3.0-3B-a800m-base — benchmark results

Provider: IBM. Released 2024-10-03. Access: Open.

Unified ELO 1308 ± 30, rank #1515 of 1776 rated models, from 38 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
EuroEval Dutch NLU - DBRD89.16Sentiment classification Score (%)63.8
EuroEval Portuguese NLU - SST-2 PT77.24Sentiment classification Score (%)51.4
EuroEval Spanish NLU - MLQA ES59.26Reading comprehension Score (%)48.2
EuroEval Italian NLU - SQuAD IT63.37Reading comprehension Score (%)44.3
EuroEval Portuguese NLU45.29NLU Average Score (%)39.4
EuroEval Spanish NLU - ScaLA ES4.28Linguistic acceptability Score (%)37.6
EuroEval Portuguese NLU - MultiWikiQA PT62.9Reading comprehension Score (%)35.9
EuroEval Portuguese NLU - HAREM36.07Named entity recognition Score (%)34.8
EuroEval Portuguese NLU - ScaLA PT4.95Linguistic acceptability Score (%)34.3
EuroEval Portuguese32.3Average Score (%)30.6
EuroEval Spanish NLU32.84NLU Average Score (%)28.3
Open LLM Leaderboard - MATH Level 54.83Score27.3

Interactive version: theaggregate.ai/model?slug=granite-3-0-3b-a800m-base · How the rankings work · Data refreshed daily, snapshot 2026-07-22.