Open Japanese LLM - Janli Exact Match: leaderboard
The "Janli Exact Match" 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 | 100 | #835 |
| 2 | llm-jp-3.1-1.8B-instruct4 | 100 | #1273 |
| 3 | Llama 3.3 70B Instruct | 91.25 | #534 |
| 4 | Llama-3.1-Tulu-3-70B | 88.61 | #482 |
| 5 | Qwen 2.5 Coder 32B Instruct | 87.08 | #528 |
| 6 | Llama 3.1 Nemotron 70B Instruct | 86.94 | #554 |
| 7 | ABEJA-Qwen2.5-32B-Japanese-v1.0 | 86.81 | #503 |
| 8 | qwen2.5-bakeneko-32B-instruct-v2 | 85.83 | #562 |
| 9 | Llama 3.1 70B Instruct | 85.56 | #561 |
| 10 | EZO-Qwen2.5-32B-Instruct | 85.42 | #471 |
| 11 | Qwen2.5-32B-Instruct-CFT | 84.44 | #375 |
| 12 | Qwen 2.5 32B Instruct | 84.31 | #498 |
| 13 | oxyge1-33B | 84.17 | #690 |
| 14 | Awqward2.5-32B-Instruct | 84.17 | #371 |
| 15 | Rombos-LLM-V2.5-Qwen-32B | 84.17 | #367 |
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-janli-exact-match · How It Works · Data refreshed daily, snapshot 2026-10-20.