Kaggle MultiLoKo - Korean: leaderboard
Metric: Exact Match (%). Source: www.kaggle.com. 48 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 57.03 |
| 2 | Gemini 2.5 Pro (Preview 06-05) | 52.8 |
| 3 | O3 | 46.8 |
| 4 | Gemini 3.5 Flash | 44.8 |
| 5 | GPT-5 | 44.4 |
| 6 | O1 | 40 |
| 7 | Gemini 2.5 Flash (Preview 05-20) | 39.6 |
| 8 | Gemini 3.1 Flash Lite (Preview) | 38 |
| 9 | Gemini 3 Flash (Preview) | 36.8 |
| 10 | Gemini 1.5 Pro (002) | 32.8 |
| 11 | GPT-5.4 | 31.2 |
| 12 | Gemini 2.0 Flash | 30.8 |
| 13 | GPT-4.1 | 27.2 |
| 14 | GPT-4o (2024-08-06) | 27.2 |
| 15 | Grok 3 | 22.8 |
Interactive version: theaggregate.ai/benchmark?slug=kaggle-multiloko-korean · How It Works · Data refreshed daily, snapshot 2026-09-19.