ECLeKTic: leaderboard
ECLeKTic is a multilingual closed-book question-answering benchmark measuring cross-lingual knowledge transfer across 12 languages, using knowledge-seeking questions drawn from Wikipedia articles that exist in only one language.
Metric: Score (imported). Source: arxiv.org. Source: benchmarklist.com. Saturation forecast: Estimated already saturated. 8 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4o | 38.8 | #343 |
| 2 | Gemini 2.0 Flash | 34.6 | #342 |
| 3 | Claude 3.5 Sonnet | 34.4 | #350 |
| 4 | Gemma 2 9B | 8.7 | #958 |
| 5 | Mistral Nemo | 7.1 | #1085 |
| 6 | Qwen 2.5 7B | 2.8 | #922 |
| 7 | OLMo 2 7B | 1.6 | #1371 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=eclektic · How It Works · Data refreshed daily, snapshot 2026-10-20.