Open Japanese LLM - Jsts Spearman: leaderboard
The "Jsts Spearman" subtask of Open Japanese LLM. A task from the Open Japanese LLM Leaderboard, evaluating models on Japanese-language knowledge, reasoning and understanding.
Metric: Score (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 418 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | llm-jp-3.1-13B-instruct4 | 90.92 | #835 |
| 2 | llm-jp-3.1-1.8B-instruct4 | 89.75 | #1273 |
| 3 | EZO-Qwen2.5-32B-Instruct | 88.23 | #471 |
| 4 | Qwen 2.5 32B | 88.1 | #546 |
| 5 | cogito-v1-preview-qwen-14B | 88.06 | |
| 6 | Qwentile2.5-32B-Instruct | 88.02 | #339 |
| 7 | Llama-3.3-Swallow-70B-Instruct-v0.4 | 87.9 | #529 |
| 8 | Awqward2.5-32B-Instruct | 87.84 | #371 |
| 9 | Qwen2.5-14B-Instruct-1M | 87.81 | #578 |
| 10 | Qwen2.5-32B-Instruct-CFT | 87.78 | #375 |
| 11 | Llama-3.1-Tulu-3-70B | 87.77 | #482 |
| 12 | Gemma 2 27B (IT) | 87.76 | #645 |
| 13 | ABEJA-Qwen2.5-32B-Japanese-v1.0 | 87.75 | #503 |
| 14 | Rombos-LLM-V2.5-Qwen-32B | 87.72 | #367 |
| 15 | Qwen 2.5 32B Instruct | 87.7 | #498 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=open-japanese-llm-jsts-spearman · How It Works · Data refreshed daily, snapshot 2026-10-20.