Nejumi 4 - jaster (0-shot) - MAWPS: leaderboard
MAWPS arithmetic word problems in Japanese, from the llm-jp-eval jaster suite as run 0-shot by Nejumi Leaderboard 4 (Weights & Biases Japan). Score is exact match of the final number (%) 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 | GPT-4.1 Mini | 99 | #363 |
| 2 | Gemma 4 31B (IT) | 99 | #178 |
| 3 | Claude 3.5 Sonnet | 99 | #362 |
| 4 | Claude Opus 5 (Max) | 99 | #16 (Claude Opus 5) |
| 5 | Gemini 3.1 Pro (Preview) | 98 | #64 |
| 6 | Gemini 3 Pro (Preview) | 98 | #71 |
| 7 | Gemini 3 Flash (Preview) | 98 | #90 |
| 8 | Gemini 3.5 Flash | 98 | #66 |
| 9 | Claude Sonnet 4.5 (Thinking) | 98 | #148 (Claude Sonnet 4.5) |
| 10 | Gemma 4 26B A4B (IT) | 98 | #293 |
| 11 | O4 Mini (2025-04-16) | 98 | #183 |
| 12 | GPT-4o (2024-11-20) | 98 | #392 |
| 13 | Kimi K2 (Thinking) | 98 | #252 (Kimi K2) |
| 14 | Qwen 3 235B A22B 2507 (Thinking) | 98 | #263 |
| 15 | O3 Mini (2025-01-31) | 98 | #249 |
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-mawps · How It Works · Data refreshed daily, snapshot 2026-11-05.