Nejumi 4 - MT-Bench (Japanese) - Roleplay: leaderboard
Japanese MT-Bench questions in the roleplay category, two-turn answers rated 1-10 by GPT-4.1 as judge in Nejumi Leaderboard 4 (Weights & Biases Japan). Nejumi stores ratings divided by 10; the board multiplies by 100 (10-100), the scale of Nejumi 4 - MT-Bench (Japanese). Higher is better.
Metric: Judge rating (1-10, x10). Source: nejumi.ai. Saturation forecast: Estimated already saturated. 145 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Kimi K3 (Max) | 100 | #35 (Kimi K3) |
| 2 | Claude Opus 5 (Max) | 100 | #16 (Claude Opus 5) |
| 3 | Gemini 3.6 Flash | 99.5 | #57 |
| 4 | GLM-5.3 (Max) | 99.5 | #44 (GLM-5.3) |
| 5 | Claude Opus 4.6 (Thinking) | 99.5 | #67 (Claude Opus 4.6) |
| 6 | Claude Opus 4.5 (Thinking) | 99.5 | #87 (Claude Opus 4.5) |
| 7 | Qwen 3.6 Max (Preview) (Thinking) | 99.5 | |
| 8 | Inkling (Max) | 99.5 | #144 (Inkling) |
| 9 | GLM-5.3 Flash (Max) | 99 | #77 (GLM-5.3 Flash) |
| 10 | Solar Pro 4 | 99 | #230 |
| 11 | Claude Sonnet 4.6 (Thinking) | 99 | #94 (Claude Sonnet 4.6) |
| 12 | Claude Opus 4.5 (Non-reasoning) | 99 | #87 (Claude Opus 4.5) |
| 13 | Claude Opus 4.8 (xHigh) | 99 | #39 (Claude Opus 4.8) |
| 14 | GPT-OSS-Swallow-120B-RL-v0.1 (Thinking) | 99 | #309 (GPT-OSS-Swallow-120B-RL-v0.1) |
| 15 | MiniMax-M3 (Thinking) | 99 | #151 (MiniMax-M3) |
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-mt-bench-japanese-roleplay · How It Works · Data refreshed daily, snapshot 2026-11-05.