I'm considering using Amazon Polly's Text-To-Speech (TTS) but want to evaluate how natural it really sounds with arbitrary examples (not the ones carefully curated by Amazon marketing folks).
All I've found is a very limited youtube video (just a few seconds in each of the languages)
Log into your AWS account, then you can experiment with different voices here: https://console.aws.amazon.com/polly/home/SynthesizeSpeech