LLM-jp4.1 Release Judge - MT-Bench (JA): leaderboard
llm-jp-judge official Sept28 release evaluation; medium reasoning effort for supported models. Exact source dated API snapshots and thinking checkpoints retained. AnswerCarefully safety behavior column deliberately not a capability board. Metric: Judge score (1–10). Higher is better. Source: https://llm-jp.nii.ac.jp/blog/llm-jp-4-1/
Metric: Judge score (1–10). Source: llm-jp.nii.ac.jp. Saturation forecast: Around January 2027. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 (2026-03-05) (Medium) | 8.87 | #57 (GPT-5.4 (2026-03-05)) |
| 2 | GPT-OSS-20B (Medium) | 7.33 | #516 (GPT-OSS-20B) |
| 3 | GPT-4o (2024-08-06) | 7.29 | #331 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #7 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=llm-jp4-1-release-judge-mt-bench-ja · How It Works · Data refreshed daily, snapshot 2026-10-17.