Icelandic LLM - WikiQA-IS: leaderboard

Metric: Score (%). Source: huggingface.co. 94 models tracked.

Top models

#ModelScore
1Claude Fable 575.39
2Gemini 3.1 Pro (Preview)67.74
3Gemini 3.7 Flash66.82
4Gemini 3 Pro (Preview)64.82
5Gemini 3.5 Flash64.68
6GPT-5.6 Sol63.21
7Gemini 3.6 Flash62.58
8Claude Opus 4.862.42
9Gemini 3 Flash (Preview)61.47
10Claude Opus 4.759.42
11GPT-5.559.18
12Gemini 2.5 Pro (Preview 03-25)52.66
13GPT-5.4 (2026-03-05)49.45
14Gemini 3.1 Flash Lite (Preview)48.18
15GPT-5.6 Luna46.87

Interactive version: theaggregate.ai/benchmark?slug=icelandic-llm-wikiqa-is · How It Works · Data refreshed daily, snapshot 2026-09-05.