Media content portions are generated based on received message inputs having words or phrases. The media content portions are recommended based on predetermined criteria, classification criteria, and/or user preferences. The media content portions are identified and extracted among media content based on predetermined criteria that can include a match of audio content with the words or phrases of the received message inputs. The media content portions correspond to the words or phrases of the message inputs and can further be recommended to a user based on additional criteria. The media content portions that are recommended can be included in multimedia message, which can be further communicated.