Open Universal Arabic ASR Leaderboard

In recent years, the enhanced capabilities of ASR models and the emergence of multi-dialect datasets have increasingly pushed Arabic ASR model development toward an all-dialect-in-one direction. This trend highlights the need for benchmarking studies that evaluate model performance in a multi-dialect setting, providing the community with insights into models' generalization capabilities.