On October 16, Yuanrong Qihang showcased its innovations at the 2025 China International New Energy and Intelligent Connected Vehicle Exhibition. During the event, the company highlighted its achievements in mass production, technological innovation, and future strategies. It also introduced the technical features and core advantages of its VLA model. As one of the first AI companies to develop the VLA model, Yuanrong Qihang integrates vision, language, and action capabilities. Compared to traditional end-to-end models, the VLA model offers powerful reasoning capabilities. It overcomes the black box problem of traditional models by linking, analyzing, and inferring causal relationships. Additionally, it naturally integrates a vast knowledge base, enhancing its ability to adapt to complex real-world driving conditions.
In August this year, Yuanrong Qixing launched the new generation of its assisted driving platform, DeepRoute IO 2.0. This platform features a self-developed VLA model. It focuses on “multi-modal + multi-chip + multi-vehicle” adaptability. It supports both LiDAR and pure vision versions. The platform can customize deployments for various mainstream passenger car models. Yuanrong Qixing has secured five fixed-point cooperation projects based on DeepRoute IO 2.0. The first batch of mass-produced vehicles will soon enter the market.
As technology evolves rapidly, Yuanrong Qihang has built a solid foundation for mass production and commercialization. In September, the company delivered over 30,000 vehicles in a single month. The total number of delivered mass-produced vehicles surpassed 130,000. This milestone marks the full acceleration of Yuanrong Qihang’s combined assisted driving solutions. By the end of 2025, over 200,000 vehicles equipped with these solutions are expected to be mass-produced and delivered.
CEO Zhou Guang of Yuanrong Qixing stated, “Yuanrong Qixing plans to launch Robotaxi services. We will base these services on mass-produced vehicles. This approach will benefit both C-end users and B-end scenarios. We aim to empower our mass production, Robotaxi, and RoadAGI businesses. This synergy will achieve general artificial intelligence in the physical world.”