: Added support for input text streaming in speech synthesis, allowing for lower latency in real-time applications [24].
: Changed the default neural voice for speech synthesis to en-US-AvaMultilingualNeural [24].
Are you writing a or technical documentation ?