SenseTime Technology has raised $500 million, setting a new record for the largest single funding round in China's domestic AI industry.

According to Dongcha Beating monitoring, Insigma Technology recently completed a new $500 million financing round, setting a new record for the largest single financing in the field of universal world models in China. This round of funding will be used for model research and technological innovation. Founded in March 2023, Insigma Technology has currently established its foundational models in three major areas: multimodal generation, real-time interaction, and physical-world action.

In the areas of multimodal generation and real-time interaction, the video generation model Vidu Q series has deeply integrated into professional content production systems such as animation, short films, advertisements, e-commerce, anime, and movies; the recently released real-time interaction model Vidu S1 has achieved voice-controlled real-time video generation, supporting real-time infinite duration, achieving 540P and 25FPS real-time interaction, with a maximum frame rate of 42FPS, and supporting custom character images and voices, enabling efficient inference on consumer-grade GPUs.

In the realm of physical-world action, Insigma Technology open-sourced the world's first "perception, prediction, and action" closed-loop universal world model Motus in December 2025, and launched the robot control model Motubrain in April 2026, achieving a 95.8% average success rate on the embodied intelligence benchmark RoboTwin 2.0, continuously advancing the universal brain of embodied intelligent robots.