Open Japanese LLM - Wiki NER SET F1: leaderboard
The "Wiki NER SET F1" 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 | 32.74 | #835 |
| 2 | llm-jp-3.1-1.8B-instruct4 | 20.35 | #1273 |
| 3 | QwQ-32B | 17.7 | #421 |
| 4 | DeepSeek-R1-Distill-Qwen-32B-Japanese | 15.93 | #638 |
| 5 | DeepSeek R1 Distill Qwen 32B | 14.16 | #655 |
| 6 | Llama 3 70B | 14.16 | #631 |
| 7 | Llama 2 70B Base | 13.27 | #1025 |
| 8 | QwQ 32B-Preview | 13.27 | #651 |
| 9 | Llama 3.1 8B Instruct | 12.39 | #1035 |
| 10 | Llama 3 70B Instruct | 12.39 | #634 |
| 11 | Llama 3.1 70B | 12.39 | #592 |
| 12 | Qwen 2.5 32B | 12.39 | #546 |
| 13 | Qwen 2.5 Coder 32B Instruct | 12.39 | #528 |
| 14 | Qwentile2.5-32B-Instruct | 12.39 | #339 |
| 15 | aya-23-35B | 11.5 | #883 |
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-wiki-ner-set-f1 · How It Works · Data refreshed daily, snapshot 2026-10-20.