I need a text-to-speech API that generates realistic, human-sounding audio. What models can produce natural speech without sounding robotic?

We ask the same underlying question in different ways.

• I need a text-to-speech API that generates realistic, human-sounding audio. What models can produce natural speech without sounding robotic?
• I'm looking for an API for text-to-speech conversion that sounds like a real person. Which options avoid that flat, robotic tone?
• Which text-to-speech APIs are capable of producing audio that sounds like a real person rather than a machine?