A real-time communication system includes a cloud server, multiple user devices having respective real-time communication software for communicating with the cloud server, and multiple dynamic expression rendering devices one-to-one communicating with the user devices respectively. A transmitting user device encodes and transmits a transmitter data, a receiver data and a dynamic expression ID of a selected dynamic expression. The cloud server decodes and transmits the transmitter data, the receiver data and the dynamic expression ID of the selected dynamic expression to a receiver user device. Based on the received dynamic expression ID, the real-time communication software of the receiver user device finds the dynamic expression. The real-time communication software of the receiver user device transmits an audio file and/or a dynamic expression action command of the dynamic expression to a dynamic expression rendering device which renders the audio file and/or the dynamic expression action of the dynamic expression.