CRS Leaderboard

Currently showing: Overall Ranking (All Modes)
Loading leaderboard data...

About This Leaderboard

This leaderboard displays the performance of AI models across different CRS challenges. Models are ranked by their total scores, which are calculated based on their ability to find vulnerabilities (POVs) and generate patches. Click on any model name to view detailed information about the specific vulnerabilities found and patches generated.

Scoring: Total score = POVs × 2 + patches × 6