Kaggle MultiLoKo - Japanese: leaderboard
Metric: Exact Match (%). Source: www.kaggle.com. 48 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 60.16 |
| 2 | Gemini 3.5 Flash | 54.8 |
| 3 | Gemini 3 Flash (Preview) | 53.6 |
| 4 | Gemini 2.5 Pro (Preview 06-05) | 46 |
| 5 | O3 | 40.8 |
| 6 | GPT-5 | 40 |
| 7 | O1 | 39.2 |
| 8 | Gemini 3.1 Flash Lite (Preview) | 36 |
| 9 | GPT-5.4 | 34.8 |
| 10 | GPT-4.1 | 34.4 |
| 11 | Gemini 2.5 Flash (Preview 05-20) | 31.2 |
| 12 | Claude Opus 4 | 30 |
| 13 | Gemini 2.0 Flash | 28.8 |
| 14 | Grok 3 | 28 |
| 15 | Claude Opus 4.1 | 27.6 |
Interactive version: theaggregate.ai/benchmark?slug=kaggle-multiloko-japanese · How It Works · Data refreshed daily, snapshot 2026-09-19.