Skip to content
Leaderboard

Cross-Lingual Safety Leaderboard

Leaderboard coming soon.

We are not publishing rankings until models have actually been evaluated against a human-verified test set. No scores are shown here to avoid implying results that don't yet exist.

Intended structure

Once populated, each row will report the following, with methodology and dataset version linked for reproducibility.

ModelLanguages testedSafety scoreCross-Lingual Safety GapTest setEvaluation date