Nejumi 4 - jaster (0-shot) - CommonsenseMoralJA: leaderboard
JCommonsenseMorality, judging whether described actions are morally wrong by Japanese commonsense, from the llm-jp-eval jaster suite as run 0-shot by Nejumi Leaderboard 4 (Weights & Biases Japan). Score is exact match (%) on Nejumi's sampled test split; higher is better.
Metric: Exact match (%). Source: nejumi.ai. Saturation forecast: Estimated already saturated. 145 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.7 (xHigh) | 99 | #51 (Claude Opus 4.7) |
| 2 | Gemini 3 Flash (Preview) | 98 | #90 |
| 3 | Gemini 3.5 Flash | 98 | #66 |
| 4 | DeepSeek V4 Flash (0731) (Max) | 98 | #84 (DeepSeek V4 Flash (0731)) |
| 5 | Gemini 2.5 Pro | 97 | #159 |
| 6 | Gemma 4 31B (IT) | 97 | #178 |
| 7 | GPT-5.6 Sol (Max) | 97 | #22 (GPT-5.6 Sol) |
| 8 | Claude Opus 4.8 (xHigh) | 97 | #39 (Claude Opus 4.8) |
| 9 | llm-jp-4.1-33B (High) | 97 | #256 (llm-jp-4.1-33B) |
| 10 | Claude Opus 4 (Non-reasoning) | 97 | #170 (Claude Opus 4) |
| 11 | DeepSeek-V4-Flash-Vision-Exp (Max) | 97 | #88 (DeepSeek-V4-Flash-Vision-Exp) |
| 12 | Gemini 3.1 Pro (Preview) | 96 | #64 |
| 13 | GPT-4.1 | 96 | #258 |
| 14 | Claude 3.5 Sonnet | 96 | #362 |
| 15 | Grok 3 | 96 | #313 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=nejumi-4-jaster-0-shot-commonsensemoralja · How It Works · Data refreshed daily, snapshot 2026-11-05.