I wish it also included latency of speech to text and text to speech APIs :)