The embodiment of the invention provides an audio processing method and device. The method comprises the steps that an audio file to be processed and a subtitle file corresponding to the audio file are acquired; the subtitle file is parsed and at least one time period is acquired, and one time period is used for describing play time of one character in the subtitle file; division processing is performed on the audio file according to the at least one time period and at least one audio data is acquired, and one time period is corresponding to one audio data; intonation information of the at least one audio data is analyzed; and an MIDI file corresponding to the audio file is generated according to the at least one time period and the intonation information of the at least one audio data. The MIDI file corresponding to the audio file can be automatically generated so that audio processing efficiency is enhanced and audio processing intelligence is enhanced.