Nejumi 4 - jaster (2-shot) - JNLI: leaderboard
JNLI (JGLUE), Japanese natural language inference, from the llm-jp-eval jaster suite as run 2-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 | Gemini 3.6 Flash | 94 | #57 |
| 2 | Claude Opus 5 (Max) | 94 | #16 (Claude Opus 5) |
| 3 | Claude Opus 4 (Non-reasoning) | 94 | #170 (Claude Opus 4) |
| 4 | Gemini 3 Flash (Preview) | 93 | #90 |
| 5 | Claude Opus 4.5 (Non-reasoning) | 93 | #87 (Claude Opus 4.5) |
| 6 | Gemini 3.1 Pro (Preview) | 92 | #64 |
| 7 | Claude Opus 4.6 (Thinking) | 92 | #67 (Claude Opus 4.6) |
| 8 | Claude Opus 4.5 (Thinking) | 92 | #87 (Claude Opus 4.5) |
| 9 | Claude Opus 4.8 (xHigh) | 92 | #39 (Claude Opus 4.8) |
| 10 | GPT-4.1 | 91 | #258 |
| 11 | Gemini 3.5 Flash | 91 | #66 |
| 12 | Gemini 3.1 Flash Lite (Preview) | 91 | #201 |
| 13 | GLM-5.3 Flash (Max) | 91 | #77 (GLM-5.3 Flash) |
| 14 | GLM-5.2 (Max) | 91 | #80 (GLM-5.2) |
| 15 | Claude Sonnet 4.6 (Thinking) | 91 | #94 (Claude Sonnet 4.6) |
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-2-shot-jnli · How It Works · Data refreshed daily, snapshot 2026-11-05.