https://medium.com/@sarty/speech-to-tex ... c58f051450
This seems to be a pretty accurate and easy way to implement speech recognition in python.
DeepSpeech also appears to be a useful option that allows for more options.
https://deepspeech.readthedocs.io/en/v0.8.2/index.html
Kitboga is using this very easy to use but simple API:-
https://github.com/wit-ai/pywit