Kaggle MultiLoKo - Spanish: leaderboard

Metric: Exact Match (%). Source: www.kaggle.com. 48 models tracked.

Top models

#ModelScore
1Gemini 3.1 Pro (Preview)70.2
2Gemini 3.5 Flash64
3Gemini 3 Flash (Preview)63.6
4Gemini 2.5 Pro (Preview 06-05)59.2
5O154.8
6O352.8
7GPT-551.6
8Gemini 2.5 Flash (Preview 05-20)51.6
9Gemini 3.1 Flash Lite (Preview)44.8
10GPT-4.144.4
11GPT-5.442.4
12GPT-4o (2024-08-06)40.8
13Claude 3.5 Sonnet (20241022)40
14O4 Mini40
15Claude 3.7 Sonnet39.6

Interactive version: theaggregate.ai/benchmark?slug=kaggle-multiloko-spanish · How It Works · Data refreshed daily, snapshot 2026-09-19.