granite-7B-instruct: benchmark results
Provider: IBM. Released 2024-05-19. Access: Open.
Aggregate Rating 1234 ± 35, rank #9380 of 10916 rated models, from 24 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard - MuSR | 40.2 | Score | 43.7 |
| Open LLM Leaderboard - GPQA | 28.52 | Score | 40.2 |
| EuroEval Portuguese NLU - MultiWikiQA PT | 52.23 | Character F1 (%) | 29.5 |
| Open LLM Leaderboard - IFEval | 29.72 | Score | 27 |
| EuroEval Norwegian Knowledge - Idioms NO | 0.38 | MCC (x100) | 26.5 |
| Open LLM Leaderboard - MMLU-Pro | 22.86 | Score | 23.5 |
| Open LLM Leaderboard - BBH | 37.23 | Score | 21.5 |
| EuroEval English Knowledge | 18.34 | Knowledge Average Score (%) | 19.1 |
| EuroEval Portuguese NLU - ScaLA PT | 0.74 | Matthews correlation coefficient (%) | 17.6 |
| EuroEval Portuguese NLU - HAREM | 26.06 | Micro F1 excluding MISC (%) | 16.6 |
| EuroEval Portuguese NLU | 27.46 | NLU Average Score (%) | 16.2 |
| Open LLM Leaderboard - MATH Level 5 | 2.04 | Score | 14.2 |
Interactive version: theaggregate.ai/model?slug=granite-7b-instruct · How It Works · Data refreshed daily, snapshot 2026-11-08.