granite-3.0-2B-base: benchmark results
IBM's Granite 3.0 2B Base, a decoder-only model trained in two stages: 10 trillion tokens from diverse domains, then 2 trillion tokens of curated high-quality data. Provider: IBM. Released 2024-10-21. Access: Open.
Aggregate Rating 1308 ± 37, rank #1463 of 1656 rated models, from 51 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| EuroEval Dutch NLU - DBRD | 89.13 | Matthews correlation coefficient (%) | 63.1 |
| EuroEval Portuguese NLU - MultiWikiQA PT | 70.02 | Character F1 (%) | 56.5 |
| EuroEval Portuguese NLU - SST-2 PT | 75.18 | Matthews correlation coefficient (%) | 45.9 |
| EuroEval Portuguese NLU | 44.99 | NLU Average Score (%) | 38.3 |
| Open LLM Leaderboard - IFEval | 38.74 | Score | 38.1 |
| Open PL LLM - PSC (multiple choice, 5-shot) | 48.81 | Binary F1 (%) | 37 |
| Open LLM Leaderboard - GPQA | 28.02 | Score | 35.2 |
| EuroEval Portuguese | 34.21 | Average Score (%) | 32.9 |
| EuroEval Danish Knowledge | 24.18 | Knowledge Average Score (%) | 31.7 |
| EuroEval Norwegian Knowledge | 6.17 | Knowledge Average Score (%) | 31.2 |
| EuroEval Norwegian Knowledge - NRK Quiz QA | 12.74 | MCC (x100) | 31.1 |
| EuroEval English Knowledge | 49.21 | Knowledge Average Score (%) | 30.3 |
Interactive version: theaggregate.ai/model?slug=granite-3-0-2b-base · How It Works · Data refreshed daily, snapshot 2026-10-20.