Nejumi 4 - GLP - Applied Language: leaderboard
An aggregate of three Nejumi 4 sub-scores: expression, judged on the MT-Bench roleplay, writing and humanities questions; translation between English and Japanese on the Asian Language Treebank corpus; and answering questions from a supplied passage in JSQuAD. The board is the unweighted mean of the three in percent, higher is better. Source: github.com/wandb/llm-leaderboard.
Metric: Score (%). Source: nejumi.ai. Saturation forecast: Estimated already saturated. 139 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.6 Flash | 90.97 | #51 |
| 2 | Claude Opus 5 | 90.8 | #12 |
| 3 | Gemini 3.5 Flash | 90.65 | #60 |
| 4 | Inkling | 90.62 | #126 |
| 5 | Kimi K3 | 90.55 | #30 |
| 6 | Hy4 preview | 90.42 | #38 |
| 7 | Gemini 3 Pro (Preview) | 90.4 | #66 |
| 8 | Claude Opus 4.7 | 90.4 | #49 |
| 9 | Qwen 3.6 Max Preview | 90.34 | #85 |
| 10 | GLM-5.3 Flash | 90.31 | #65 |
| 11 | Inkling Small | 89.99 | #138 |
| 12 | Claude Opus 4.6 | 89.94 | #63 |
| 13 | GPT-5.4 (2026-03-05) | 89.94 | #59 |
| 14 | Nex N2 Pro | 89.89 | #107 |
| 15 | Gemini 3 Flash (Preview) | 89.88 | #82 |
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-glp-applied-language · How It Works · Data refreshed daily, snapshot 2026-10-20.