granite-3.1-8B-base Benchmark Results

granite-3.1-8B-base: benchmark results

Provider: IBM. Released 2024-12-18. Access: Open.

Aggregate Rating 1412 ± 32, rank #5544 of 10916 rated models, from 56 benchmark results.

Benchmark	Score	Metric	Percentile
Open LLM Leaderboard - GPQA	32.13	Score	76
Open Japanese LLM - Wiki NER SET F1	8.85	Score (%)	72
Open LLM Leaderboard - MATH Level 5	9.44	Score	45.5
Open LLM Leaderboard - MMLU-Pro	32.32	Score	44.4
Open Japanese LLM - Xlsum JA Bert Score JA F1	68.63	Score (%)	44.2
Open LLM Leaderboard - IFEval	42.21	Score	43.5
Open Japanese LLM - SUM	9.21	Score (%)	43
Open Japanese LLM - Xlsum JA Rouge2	9.21	Score (%)	43
Open Japanese LLM - Xlsum JA Rouge1	24.29	Score (%)	41.1
Open LLM Leaderboard - BBH	47.77	Score	41
Open Japanese LLM - Xlsum JA RougeLsum	20.74	Score (%)	40.9
Open Japanese LLM - Jsem Exact Match	74.43	Score (%)	38

Interactive version: theaggregate.ai/model?slug=granite-3-1-8b-base · How It Works · Data refreshed daily, snapshot 2026-11-08.