reasoning benchmark
ECLeKTic
A multilingual closed-book question answering dataset that evaluates cross-lingual knowledge transfer in large language models across 12 languages, using knowledge-seeking questions based on Wikipedia articles that exist only in one language
Updated Aug 7, 2026
Published models8
Registry coverage8
MetricScore
EvidenceB
ECLeKTic leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | GO | 19.0% | 100.0% | 8 | C | |
| 2 | GO | 16.7% | 85.7% | 8 | C | |
| 3 | GO | 10.3% | 71.4% | 8 | C | |
| 4 | GO | 4.6% | 57.1% | 8 | C | |
| 5 | GO | 2.5% | 42.9% | 8 | C | |
| 6 | GO | 2.5% | 28.6% | 8 | C | |
| 7 | GO | 1.9% | 14.3% | 8 | C | |
| 8 | GO | 1.4% | 0.0% | 8 | C |
Score distribution
Top published rows on the benchmark's original scale.
ECLeKTic highlights
The top published results on this benchmark's own scale.
What is ECLeKTic?
Definition and scoring fields from the benchmark registry.
A multilingual closed-book question answering dataset that evaluates cross-lingual knowledge transfer in large language models across 12 languages, using knowledge-seeking questions based on Wikipedia articles that exist only in one language
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- ECLeKTic
- Modality
- text
- Primary category
- reasoning
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- eclektic|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about ECLeKTic.
Which model scores highest on ECLeKTic?
Gemma 3n E4B Instructed is currently ranked first with 19.0%.
What does ECLeKTic measure?
A multilingual closed-book question answering dataset that evaluates cross-lingual knowledge transfer in large language models across 12 languages, using knowledge-seeking questions based on Wikipedia articles that exist only in one language
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
8 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is preserved as source-native evidence but is not eligible for the current overall score.