I have seen 2 deficiencies in most text-to-speech applications:
1. Premium sound may not be available, most of them have standard AI sounds.
2. Most of them can convert small text such as 5000–7000 characters at one time.
In this new application, you can choose one of the WaveNet voices to get better quality sound and you can convert up to 140000 characters.
Text to Speech Application Application address: https://text2speech.sysarticles.com/
For more information and samples about WaveNet: https://deepmind.com/blog/article/wavenet-generative-model-raw-audio