A method and system for generating multi-modal callback messages for a user of a wireless terminal connected to an access network. A multi-modal response is generated in response to a request for information with a multi-modal callback server. The multi-modal response is transmitted to the wireless terminal using the access network. A text-based response is included in the multi-modal response that includes a means for allowing a predetermined portion of the multi-modal response to be read aloud to the user of the wireless terminal or another message may be read aloud to the user.