The service provides a hosted version of the provider's installed text-to-speech (TTS) conversion software. It generates spoken language from input text in a variety of languages using a selection of speaking voices. Requests can fine-tune aspects of the audio output such as pitch and pacing to adjust for specific needs.
API methods support either "simple speak" conversion of a basic text selection or "extended speak" conversion with modifications to default pitch, rate of speech, and pronunciation via SSML markup. Methods also support management of multiple lexicons and abbreviation sets to control text conversions for specific environments. Tags allow influence over emotional tone and vocal gestures.