About us

MOSI Intelligence (full name: Shanghai MOSI Intelligent Technology Co., Ltd.), headquartered in Shanghai, is a large model innovation enterprise for the next generation of situational interactive intelligence. It was incubated in the OpenMOSS team of Shanghai Chuangzhi College and Fudan University, with Professor Qiu Xipeng as the chief scientist. With multimodal large models as the core engine, it promotes the large-scale implementation of intelligent interaction, embodied intelligence, and related industrial scenarios.

The company is committed to the long-term direction of "multimodal situational intelligence" and the current focus of "real-time interaction system", aiming to connect AI, humans, and the physical world, and create a real-time interaction system and infrastructure for physical world AI, allowing intelligence to step out of dialogue boxes and enter the real world, continuously perceive situations, interact naturally, and take action. The relevant model capabilities have achieved open source leading results in multiple directions such as audio understanding, speech synthesis, real-time visual understanding, audio and video generation, and real-time interaction.

MOSI combines strong academic expertise with deep industry experience. Around 50% of our team hold doctorates. Core members come from the Shanghai Innovation Institute and Fudan University, joined by experienced professionals from leading companies. Together, they span academic research, product engineering, and commercial operations.