In a multimedia interactive telephony system, a voice service server generates dynamic content intended for consumption by a communication device. The dynamic content is sent to a gateway where it is transformed from to an intermediate content format appropriate for rendering at the communication device. The user may interact with the transformed dynamic content rendered on the communication device, causing the arguments to be sent to the voice server, thus allowing user interactivity with the voice service. The voice services server may also generate dynamic content for simultaneous consumption by multiple communication devices, each of which may independently render an intermediate content format appropriate to it. The voice services server may also generate the dynamic content for the communication device while the communication device is not currently engaged in an active call.