Gasgoo Munich-Lightwheel AI, a rising "unicorn" in embodied AI data, has recently secured 1 billion yuan in a combined A++ and A+++ funding round, according to Gasgoo. The round drew in a mix of industrial players and financial backers, including New Hope Group, Dingbang Investment (the family office of San'an Optoelectronics' chairman), Aux, and Dingshi Asset Management. Financial investors such as Jiantou Huake, Guofang Innovation, Daohe Long-term Investment, and Fresh Capital also participated.
The company plans to channel the capital into three key areas: continued R&D for its physics simulation engine, upgrades to its large-scale model evaluation system, and the expansion of global delivery and local deployment capabilities. These moves aim to solidify Guaigu's leadership in physical AI data and simulation infrastructure.
Founded in 2023, the startup is led by Xie Chen, a globally recognized simulation expert who previously headed autonomous driving simulation at Nvidia, Cruise, and Nio. Xie brings deep experience in developing and deploying synthetic data from the ground up.
Image Source: Lightwheel AI
At its core, Lightwheel AI builds the simulation and data infrastructure that powers the physical AI ecosystem. As robots move into real-world environments, the company argues, the true bottleneck is no longer hardware—it's the ability to scale data.
To address this, the startup has built an embodied data and simulation engine centered on a three-layer architecture: World, Behavior, and Eval. This system spans the entire chain—from physically realistic simulation and large-scale data production to model capability evaluation.
Image Source: Lightwheel AI
World Layer: Built around a proprietary tech stack unifying solving, measurement, and generation, this layer features a self-developed solver capable of high-precision, real-time processing for rigid bodies, soft bodies, and fluids. Combined with its pioneering "physical measurement factory" and virtual-real alignment methodology, alongside large-scale non-rigid asset production, Guaigu has created a closed loop that moves seamlessly from physical accuracy to data scaling.
Behavior Layer: Guaigu has developed the world's largest non-ego data engine, covering both synthetic simulation data and human video data (EgoSuite) to achieve mass production of embodied data. Its solutions are being delivered globally, with production capacity outpacing the industry, providing high-frequency, scalable fuel for embodied foundation models.
Eval Layer: The company launched RoboFinals, the industry's first simulation evaluation platform that meets industrial-grade standards, supports frontier large models, and offers sufficient difficulty. RoboFinals has set the benchmark for embodied AI evaluation, driving the large-scale adoption of industry assessment ecosystems.
On the commercial front, Lightwheel AI is already delivering globally across three key areas: synthetic simulation data, simulation evaluation, and human video data. Official figures show revenue surged tenfold in 2025. Even more striking, projected revenue for the first quarter of 2026 is set to surpass the company's total earnings for all of 2025.
Guaigu's partner roster includes heavyweights like Nvidia, Google, Figure AI, 1X Technologies, ByteDance, Alibaba, and Zhiyuan. Among the world's leading embodied AI teams, sources indicate that over 80% of simulation assets and synthetic data originate from Lightwheel AI.
About "Seeds Discovery":
Gasgoo's "Seeds Discovery" column is designed to connect startups with ecosystem partners, investors, and local governments, empowering the broader supply chain. Since its inception, the column has focused on uncovering the companies, technologies, and business models that are shaping the intelligent transformation, helping to drive industrial innovation. Gasgoo data shows that nearly every startup featured in "Seeds Discovery" has successfully secured resources within the industry ecosystem.
