llm-jp-3.1-13B: benchmark results
LLM-jp 3.1 pretrained release developed by NII's LLM-jp project, using Llama architecture and a 4,096-token context window. Provider: National Institute of Informatics. Access: Open.
Aggregate Rating 1541 ± 32, rank #1823 of 10916 rated models, from 52 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Japanese LLM - RC | 93.07 | Score (%) | 99.8 |
| Open Japanese LLM - Jamp Exact Match | 83.62 | Score (%) | 98.8 |
| Open Japanese LLM - Jsem Exact Match | 83.14 | Score (%) | 98.8 |
| Open Japanese LLM - Jsick Exact Match | 88.47 | Score (%) | 98.8 |
| Open Japanese LLM - FA | 40.26 | Score (%) | 98.6 |
| Open Japanese LLM - Janli Exact Match | 98.06 | Score (%) | 98.6 |
| Open Japanese LLM - Jsts Spearman | 89.11 | Score (%) | 98.6 |
| Open Japanese LLM - Wiki Dependency SET F1 | 70.93 | Score (%) | 98.6 |
| Open Japanese LLM - Wiki PAS SET F1 | 23.8 | Score (%) | 98.6 |
| Open Japanese LLM - Jsts Pearson | 91.96 | Score (%) | 98.3 |
| Open Japanese LLM - NLI | 84.99 | Score (%) | 98.3 |
| Open Japanese LLM - Wikicorpus J TO E Bert Score EN F1 | 91.62 | Score (%) | 96 |
Interactive version: theaggregate.ai/model?slug=llm-jp-3-1-13b · How It Works · Data refreshed daily, snapshot 2026-11-08.