RakutenAI-3.0: benchmark results
Provider: Other.
Aggregate Rating 1585 ± 27, rank #1183 of 10916 rated models, from 96 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Nejumi 4 - BFCL - Non-Live AST | 87.04 | Accuracy (%) | 100 |
| Nejumi 4 - MT-Bench (Japanese) - Math | 100 | Judge rating (1-10, x10) | 89.7 |
| Nejumi 4 - MT-Bench (Japanese) - Roleplay | 98.5 | Judge rating (1-10, x10) | 84.5 |
| Nejumi 4 - jaster (2-shot) - JSQuAD | 80.82 | Character F1 (x100) | 82.1 |
| Nejumi 4 - GLP - Applied Language | 89.22 | Score (%) | 72.3 |
| Nejumi 4 - Toxicity - Social Norms | 97.22 | Criteria met (%) | 71.7 |
| Nejumi 4 - GLP - Information Retrieval | 79.85 | Score (%) | 71.5 |
| Nejumi 4 - BFCL - Relevance Detection | 77.78 | Accuracy (%) | 70.7 |
| Nejumi 4 - jaster (0-shot) - JSQuAD | 78.89 | Character F1 (x100) | 65.5 |
| Nejumi 4 - BFCL - Live AST | 70.37 | Accuracy (%) | 59 |
| Nejumi 4 - jaster (2-shot) - JSeM | 78 | Exact match (%) | 57.9 |
| Nejumi 4 - GLP - Expression | 97 | Score (%) | 57.7 |
Interactive version: theaggregate.ai/model?slug=rakutenai-3-0 · How It Works · Data refreshed daily, snapshot 2026-11-08.