Kaggle MultiLoKo - Bengali: leaderboard
Metric: Exact Match (%). Source: www.kaggle.com. 48 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 45.2 |
| 2 | Gemini 2.5 Pro (Preview 06-05) | 42.8 |
| 3 | Gemini 3.5 Flash | 42.62 |
| 4 | Gemini 3 Flash (Preview) | 37.2 |
| 5 | O3 | 36.8 |
| 6 | Gemini 2.5 Flash (Preview 05-20) | 36.8 |
| 7 | Gemini 3.1 Flash Lite (Preview) | 32.8 |
| 8 | GPT-4.1 | 32.4 |
| 9 | GPT-4o (2024-08-06) | 32.4 |
| 10 | GPT-5 | 32 |
| 11 | Claude Sonnet 4 | 30.4 |
| 12 | Gemini 2.0 Flash | 29.6 |
| 13 | O1 | 29.2 |
| 14 | Claude Opus 4.1 | 29.2 |
| 15 | Claude Opus 4.8 | 28.8 |
Interactive version: theaggregate.ai/benchmark?slug=kaggle-multiloko-bengali · How It Works · Data refreshed daily, snapshot 2026-09-19.