EuroEval Icelandic Common Sense Reasoning — leaderboard

Metric: Common Sense Reasoning Average Score (%). Source: euroeval.com. 324 models tracked.

Top models

#ModelScore
1GPT-595.1
2Gemini 2.5 Pro92.47
3Gemini 3 Pro (Preview)89.84
4O3 (2025-04-16)89.24
5GPT-5 (High)89.2
6Gemini 2.5 Pro (Preview 03-25)88.43
7Claude Sonnet 4.5 (Thinking)88.1
8Gemini 2.5 Flash (Preview 04-17)86.14
9GPT-5 Mini (High)85.66
10Gemini 2.5 Flash (Thinking)81.9
11Grok 4.1 Fast (Reasoning)81.87
12GPT-5 Mini80.19
13GPT-5.4 Mini (High)78.79
14Claude 3.7 Sonnet (20250219) (Thinking)76.58
15Grok 3 Mini76.18

Interactive version: theaggregate.ai/benchmark?slug=euroeval-icelandic-common-sense-reasoning · How the rankings work · Data refreshed daily, snapshot 2026-07-22.