llm-jp-4.1-32B-a3B (Thinking): benchmark results
Provider: Other. Released 2026-09-28. Access: Open.
Aggregate Rating 1555 ± 21, rank #851 of 2159 rated models, from 38 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Japanese LLM v2 - M-IFEval-Ja | 56.4 | Prompt-Level Strict Accuracy (%) | 100 |
| Open Japanese LLM v2 - JCoLA Out-of-Domain | 81.35 | Accuracy (%) | 86.4 |
| Open Japanese LLM v2 - JFinQA | 65.6 | Accuracy (%) | 80 |
| Open Japanese LLM v2 - JCulture-MCQ | 52.24 | Set F1 (%) | 77.8 |
| Open Japanese LLM v2 - JamC-QA | 52.75 | Exact Match (%) | 65.2 |
| Open Japanese LLM v2 - JCoLA In-Domain | 76.88 | Accuracy (%) | 63.6 |
| Nejumi 4 - ALT - Bias | 87.1 | Score (%) | 56.3 |
| Nejumi 4 - ALT - Controllability | 87.11 | Score (%) | 52.2 |
| Nejumi 4 - GLP - Information Retrieval | 77.62 | Score (%) | 47.8 |
| Nejumi 4 - jaster (0-shot) | 79.62 | Score (%) | 41.8 |
| Open Japanese LLM v2 - JBLiMP | 79.15 | Accuracy (%) | 36.4 |
| Nejumi 4 - GLP - Applied Language | 87.46 | Score (%) | 34.3 |
Interactive version: theaggregate.ai/model?slug=llm-jp-4-1-32b-a3b-thinking · How It Works · Data refreshed daily, snapshot 2026-10-18.