Open Japanese LLM Leaderboard — leaderboard

Japanese language evaluation across 43 tasks including JCommonsenseQA, JMMLU, JSQuAD, MAWPS, and multilingual benchmarks.

Metric: Average Score (%). Source: huggingface.co. Status: saturation imminent. 418 models tracked.

Top models

#ModelScore
1QwQ 32B-Preview67.15
2Awqward2.5-32B-Instruct66.65
3Llama 3.1 70B Instruct66.38
4Llama 3.1 Nemotron 70B Instruct66.24
5Qwen2.5-32B-Instruct-abliterated-v265.61
6Qwen 2.5 32B Instruct65.53
7Qwen2.5-32B-Instruct-CFT65.51
8Saka-14B65.17
9tempmotacilla-cinerea-030864.65
10Qwentile2.5-32B-Instruct64.57
11Qwen2.5-14B-Instruct-1M64.53
12Qwen 2.5 14B64
13oxyge1-33B63.14
14Qwen 2.5 14B Instruct62.85
15Llama 3 70B62.73

Interactive version: theaggregate.ai/benchmark?slug=open-japanese-llm-leaderboard · How the rankings work · Data refreshed daily, snapshot 2026-07-22.