stt-server Problem Why? Google Speech Recognition is already a RESTful API. However, it is limited to input Lossless audio file. If you have a .mp3, or .aac file, this server use ffmpeg to convert the audio file into FLAC format and send it to Google API.