reasoning benchmark
MEGA XStoryCloze
XStoryCloze as part of the MEGA benchmark suite. A cross-lingual story completion task that consists of professionally translated versions of the English StoryCloze dataset to 10 non-English languages. Requires models to predict the correct ending for a given four-sentence story, evaluating commonsense reasoning and narrative understanding.
Updated Aug 7, 2026
Published models2
Registry coverage2
MetricScore
EvidenceB
MEGA XStoryCloze leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | MI | 82.8% | 100.0% | 2 | C | |
| 2 | MI | 73.5% | 0.0% | 2 | C |
Score distribution
Top published rows on the benchmark's original scale.
MEGA XStoryCloze highlights
The top published results on this benchmark's own scale.
What is MEGA XStoryCloze?
Definition and scoring fields from the benchmark registry.
XStoryCloze as part of the MEGA benchmark suite. A cross-lingual story completion task that consists of professionally translated versions of the English StoryCloze dataset to 10 non-English languages. Requires models to predict the correct ending for a given four-sentence story, evaluating commonsense reasoning and narrative understanding.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- MEGA XStoryCloze
- Modality
- text
- Primary category
- reasoning
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- mega-xstorycloze|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about MEGA XStoryCloze.
Which model scores highest on MEGA XStoryCloze?
Phi-3.5-MoE-instruct is currently ranked first with 82.8%.
What does MEGA XStoryCloze measure?
XStoryCloze as part of the MEGA benchmark suite. A cross-lingual story completion task that consists of professionally translated versions of the English StoryCloze dataset to 10 non-English languages. Requires models to predict the correct ending for a given four-sentence story, evaluating commonsense reasoning and narrative understanding.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
2 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.