WeLion New Energy has branched out to establish an AI company, using silicon-based computing power to accelerate solid-state battery materials R&D, control trial-and-error costs, and build a cloud-based data moat.
In mid-2026, on the eve of domestic solid-state battery production scaling to an industrial level, the interplay between new-energy unicorns and cutting-edge technology is evolving toward deeper underlying logic. Recently, an entity named Vota (Beijing) Artificial Intelligence Technology Co., Ltd. quietly came into being, drawing intense attention from the hard-tech investment community. According to the latest business registration information disclosed by Tianyancha App, the company has a registered capital of just one million yuan, with Yang Youwen listed as its legal representative. Its business scope is tightly focused on soft-tech strongholds including AI industry application system integration services, AI public data platforms, and AI general application systems.
The superficially popular consensus within the industry is that this is merely a reactive, cross-sector gesture by Weilan New Energy to follow trends outside its core solid-state battery business, or a light-asset PR move to ride the AI large-model wave. Such facile commentary seriously underestimates the brutal material bottlenecks confronting the solid-state battery industry at this technological inflection point. As a unicorn deeply tied to leading automakers like NIO and renowned for its semi-solid and all-solid-state battery R&D, Weilan New Energy's strategic placement of an AI entity at this moment has a deeper underlying cause: it is a hard-core maneuver by a new-energy hardware giant to overcome core pain points—such as excessively long new-material development cycles and insufficient BMS data control under high-discharge systems—by substituting "silicon-based computing power" for "traditional trial-and-error materials."
We must clearly recognize that all-solid-state battery R&D is, at its core, a particle-level contest over electrolyte interface stability. The traditional laboratory craftsmanship model relies on researchers repeatedly swapping formulations and running non-standard physical and chemical experiments, with each ratio test incurring prolonged time losses and hefty material amortization. The introduction of AI industry application system integration and general application systems harbors a hidden commercial agenda: the full-scale adoption of materials informatics. By leveraging the computational power of large models, tens of thousands of electrolyte materials' molecular structures and electrochemical reaction pathways can be simulated instantaneously in a virtual sandbox, compressing what once required years or even decades of R&D pipelines into mere days. This constitutes an absolute dimensionality reduction in efficiency.
Looking through the shareholding intersections revealed by Tianyancha, Vota AI is co-owned by Yang Youwen, Beijing Vota Technology Partnership (Limited Partnership), and Beijing Weilan New Energy Technology Co., Ltd. Within this carefully engineered interest chain, the one-million-yuan registered capital may seem as thin as paper, yet it reflects the seasoned acumen of hard-tech veterans in asset planning. As an agile outpost established outside Weilan's massive heavy-asset parent body, this independent legal entity shell forms a perfect financial and compliance risk isolation barrier. The fusion of new materials and AI is fraught with enormous uncertainty over technological misdirection; strictly locking all sunk costs of early-stage trial and error within the physical boundary of that one million yuan not only allows for absorbing top-tier algorithmic talent with minimal compliance friction but also prevents financial losses from exploratory phases from contaminating the parent company's balance sheet.
The more critical long-tail monetization logic lies in the "public data platform" prominently listed in the business scope. As smart electric vehicles enter the second half of the stock-market attrition war, battery health management and thermal runaway prediction have shifted from single-vehicle hardware competition into a battle for cloud-based data sovereignty. Through a wholly-owned or partially-held AI entity, Weilan New Energy is attempting to build a full-stack closed loop from the physical cell layer to the cloud-based AI prediction layer, using algorithms to lock in high-stickiness long-tail traffic and establishing a technological moat that shuts out traditional component suppliers.
The evolution of commerce has always been ruthlessly unforgiving. This latest Beijing coordinate recorded in the Tianyancha archive clearly proclaims that the energy giants of the future will no longer be defined solely by gigawatt-hours of production capacity. Whoever can first stitch together cold silicon-based computing power with the muddy material workshop in mutual benefit will be the one to secure an exemption ticket in the next round of cyclical upheaval.
