hei margu,
sounds like you mean “speech-to-text” (not text-to-speech), right?
i’m not aware of anything readily available other than the following works-in-progress:
- an offline version: VL.Speech
- requires internet connection: Azure Speech Recognizer Demo