I want to use google speech api to convert audio files to string. But it seems that it only accepts .raw files. >Solution : You don’t have to convert them. See the Introduction to audio encoding documentation which discusses file formats vs encoding and shows the Supported audio encodings.