Nejumi 4 - MT-Bench (Japanese) - Coding: leaderboard
Japanese MT-Bench questions in the coding 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 | Gemini 2.5 Pro | 100 | #159 |
| 2 | Gemma 4 31B (IT) | 100 | #178 |
| 3 | Gemini 3 Flash (Preview) | 100 | #90 |
| 4 | Gemini 3.1 Flash Lite (Preview) | 100 | #201 |
| 5 | Kimi K3 (Max) | 100 | #35 (Kimi K3) |
| 6 | Kimi K2.5 (Thinking) | 100 | #152 (Kimi K2.5) |
| 7 | GLM-5.3 (Max) | 100 | #44 (GLM-5.3) |
| 8 | Claude Opus 5 (Max) | 100 | #16 (Claude Opus 5) |
| 9 | GLM-5.3 Flash (Max) | 100 | #77 (GLM-5.3 Flash) |
| 10 | DeepSeek V4 Pro (0813) (Max) | 100 | #58 (DeepSeek V4 Pro (0813)) |
| 11 | O3 (2025-04-16) (High) | 100 | #125 (O3 (2025-04-16)) |
| 12 | mistral-large-4-oss (High) | 100 | #60 (mistral-large-4-oss) |
| 13 | Muse Spark Open (xHigh) | 100 | #42 (Muse Spark Open) |
| 14 | Qwen 3.5 27B (Thinking) | 100 | #192 (Qwen 3.5 27B) |
| 15 | Claude Opus 4.7 (xHigh) | 100 | #51 (Claude Opus 4.7) |
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-coding · How It Works · Data refreshed daily, snapshot 2026-11-05.