en.Wedoany.com Reported - On July 2, Chinese general-purpose artificial intelligence company VAST (Sanqi Wanwu) announced the completion of a strategic A3 round financing of over 1 billion yuan. Following this round, the company will continue to invest in the development of AI 3D large models and world models.

The financing lineup covers industrial capital, game developers, internet strategic investors, local industrial funds, and financial investment institutions. Participants include industrial players such as China Geely Capital; game developers such as China 4399, China Tangu Games, and China Giant Network; a well-known internet strategic investor; as well as investment institutions and local industrial funds including China E-Town Capital, China CPE Capital, China Yida Capital, China Chengdu Science and Technology Innovation Investment, China Fosun Capital, China Huakong Fund, China Wuxi Venture Capital, China Muhua Science and Technology Innovation, and China Ceyuan Capital. Existing shareholders China Ince Capital and China Yunsheng Capital continued to follow the investment. For China VAST, the A3 round is not merely about supplementing working capital but rather doubling down on AI 3D generation, world models, spatial content creation, and global commercialization. Previously disclosed information shows that China VAST completed nearly $200 million in Series A financing this year, with funds also directed toward AI 3D large models, general-purpose world models, algorithm iteration, data accumulation, and global market expansion.

China VAST's business focus is advancing AI generation capabilities from text, images, and video to 3D spatial content. Its core directions include AI 3D large models, world models, a UGC interactive content platform, and 3D content generation services for developers, creators, and industrial clients. Unlike traditional 3D modeling, which relies on professional software and manual production, AI 3D large models aim to generate editable, usable, and scene-ready 3D assets through text, images, or multimodal inputs. Previously disclosed materials indicate that China VAST has developed its own foundational 3D large model, with ecosystem collaborations covering enterprises and developers. Its Tripo Studio platform has already attracted a large number of creators and generated numerous 3D models. This is also a key reason game developers participated in this financing round, as game development, virtual scenes, character assets, interactive content, and UGC gameplay all require lower-cost, higher-efficiency 3D content production tools.

World models are another main line of China VAST's subsequent investment. The value of world models lies not only in generating visually appealing 3D models but also in enabling the system to understand and infer spatial relationships, objects, actions, state changes, and interactions. China VAST's previously disclosed Project Eden emphasizes decoupling underlying state inference from visual presentation, meaning the model must not only generate images but also understand how objects in a scene change, interact, and remain consistent across continuous states. For gaming, film and television, digital twins, robot simulation, industrial design, and autonomous driving simulation, world models could become a foundational capability. The participation of game developers in this financing indicates that AI 3D and world models are no longer just laboratory technologies but are beginning to integrate with content production, interactive entertainment, and commercial scenarios.

Following the completion of this financing round, China VAST will continue to invest in core algorithm iteration, data accumulation, and top-tier talent acquisition, while advancing global market commercialization and industrial ecosystem development. For AI 3D large models to truly form product capabilities, they require large amounts of high-quality 3D data, stable generation pipelines, controllable editing capabilities, cross-platform format adaptation, developer tools, and commercial distribution channels. World models further require higher-level data organization, physical consistency, temporal inference, and multi-scenario generalization capabilities. For industrial clients, whether a model can generate assets directly usable in production workflows is more important than mere demonstration effects. China VAST's consecutive large-scale financing rounds indicate that capital is focusing on the potential of AI 3D content production transitioning from tool-based experiences to industrial-level applications, and also suggest that 3D asset generation, interactive content platforms, and spatial intelligence are becoming important directions in the application layer of general-purpose artificial intelligence.