Here's a sample wav output from using their swagger endpoint: https://drive.google.com/file/d/15y83NSXOCrEW9v9eQVCy6oHcWJ8...

Why does the voice/pronunciation have such drastic volume spikes and dips?