Kaggle MultiLoKo - Farsi: leaderboard
Metric: Exact Match (%). Source: www.kaggle.com. 48 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 47.54 |
| 2 | Gemini 3.5 Flash | 42.4 |
| 3 | Gemini 3 Flash (Preview) | 36.8 |
| 4 | Gemini 2.5 Pro (Preview 06-05) | 27.6 |
| 5 | Gemini 3.1 Flash Lite (Preview) | 26 |
| 6 | GPT-5 | 25.2 |
| 7 | Gemini 2.5 Flash (Preview 05-20) | 24.8 |
| 8 | O1 | 24.4 |
| 9 | Gemini 2.0 Flash | 24 |
| 10 | O3 | 24 |
| 11 | Gemini 1.5 Pro (002) | 23.2 |
| 12 | GPT-5.4 | 20.8 |
| 13 | Gemini 1.5 Flash (002) | 16.4 |
| 14 | GPT-4o (2024-08-06) | 16.4 |
| 15 | O4 Mini | 16.4 |
Interactive version: theaggregate.ai/benchmark?slug=kaggle-multiloko-farsi · How It Works · Data refreshed daily, snapshot 2026-09-19.