Today, we announce the second release of PazaBench, our benchmark for evaluating Automatic Speech Recognition (ASR) models across African languages.