A multi-language benchmark designed to evaluate AI coding agents across diverse programming tasks and languages. It contains over 2,000 curated issues from 21 real-world repositories, covering Java, ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results