SpeechSpeech - Speech is text-to-speech produced by the API. This will have a voice associated with it as well - such as Joanna or Brian API endpoints allow you to create text-to-speech (audio files) on API.audio. Speech audio files are created from Script objects (see ScriptScript - A script is our 'unit' of audio. Consider this like your story. Scripts can have many sections within them. API), or instantaneously through /speech/sync endpoint.
When creating Speech for Script, each Script section is processed separately, which allows for separate voice and Speech parameters per section.