With the gradual progress of virtual reality technology, it is increasingly difficult for traditional video conferencing to meet people’s needs for remote collaboration. In view of the shortcomings of traditional video conferencing system in spatial presence and cross-language collaboration, this paper presents an intelligent virtual conference system based on virtual reality which supports multi-language real-time interaction. The Photon server is used in the system to realize efficient real-time communication among multiple people. The spatial voice-title synchronization mechanism is innovatively constructed to realize the dynamic anchoring of the translated text to the speaker’s head area. Developed cross-platform file sharing function, support PPT/Word document real-time upload, synchronous display and remote control. In this work, the system usability scale was used to test the system. 8 test participants all recognized that the system improved the efficiency of remote assistance, and believed that this work was of great help to the immersion and fluency of cross-language meetings.