Nejumi 4 - jaster (0-shot) - JaMP: leaderboard
JaMP, Japanese natural language inference on temporal reasoning, 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: Around April 2027. 145 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.8 (xHigh) | 85 | #39 (Claude Opus 4.8) |
| 2 | Claude Opus 5 (Max) | 84 | #16 (Claude Opus 5) |
| 3 | GPT-5 (2025-08-07) (High) | 81 | |
| 4 | O3 (2025-04-16) (High) | 81 | #125 (O3 (2025-04-16)) |
| 5 | O4 Mini (2025-04-16) | 80 | #183 |
| 6 | Claude Opus 4.6 (Thinking) | 80 | #67 (Claude Opus 4.6) |
| 7 | GPT-5.6 Luna (Max) | 79 | #79 (GPT-5.6 Luna) |
| 8 | GPT-5.6 Terra (Max) | 79 | #40 (GPT-5.6 Terra) |
| 9 | GPT-5.6 Sol (Max) | 78 | #22 (GPT-5.6 Sol) |
| 10 | Kimi K3 (Max) | 78 | #35 (Kimi K3) |
| 11 | Qwen 3.5 122B A10B (Thinking) | 78 | #184 (Qwen 3.5 122B A10B) |
| 12 | Qwen 3.6 Max (Preview) (Thinking) | 78 | |
| 13 | GPT-5.1 (2025-11-13) (High) | 78 | #107 (GPT-5.1 (2025-11-13)) |
| 14 | Kimi K3 (Thinking) | 78 | #35 (Kimi K3) |
| 15 | Qwen 3.8 27B (xHigh) | 77 | #123 (Qwen 3.8 27B) |
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-jamp · How It Works · Data refreshed daily, snapshot 2026-11-05.