Open Japanese LLM - Jsts Pearson: leaderboard
The "Jsts Pearson" 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 | 93.75 | #835 |
| 2 | llm-jp-3.1-1.8B-instruct4 | 92.67 | #1273 |
| 3 | Qwen 2.5 32B | 90.84 | #546 |
| 4 | cogito-v1-preview-qwen-14B | 90.8 | |
| 5 | Llama-3.3-Swallow-70B-Instruct-v0.4 | 90.7 | #529 |
| 6 | ABEJA-Qwen2.5-32B-Japanese-v1.0 | 90.42 | #503 |
| 7 | DeepSeek-R1-Distill-Qwen-32B-Japanese | 90.31 | #638 |
| 8 | Qwentile2.5-32B-Instruct | 90.3 | #339 |
| 9 | Llama 3.1 70B | 90.18 | #592 |
| 10 | DeepSeek R1 Distill Qwen 32B | 90.15 | #655 |
| 11 | Llama-3.1-Tulu-3-70B | 90.15 | #482 |
| 12 | EZO-Qwen2.5-32B-Instruct | 90.14 | #471 |
| 13 | Lamarckvergence-14B | 90.11 | #545 |
| 14 | Gemma 2 27B (IT) | 90.06 | #645 |
| 15 | Qwen 2.5 14B | 89.94 | #725 |
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-pearson · How It Works · Data refreshed daily, snapshot 2026-10-20.