Spirit AI Secures Nearly ¥2 Billion in Consecutive Funding Rounds as Embodied Intelligence Race Heats Up

The embodied AI sector in China is witnessing a surge unlike anything the country's tech ecosystem has seen in recent years. Spirit AI, a Beijing-based robotics and artificial intelligence startup, has officially announced the successful closure of two back-to-back funding rounds that together amount to nearly 2 billion yuan — approximately $275 million USD. This remarkable achievement in AI funding places Spirit AI firmly among the most well-capitalized players in the global embodied intelligence race. The sheer speed at which these consecutive rounds were closed is itself a testament to the growing confidence that both private and institutional investors have in the transformative potential of embodied AI systems. For anyone tracking AI funding news coming out of China, this development stands as one of the most significant financing events of early 2026, setting a powerful precedent for how quickly next-generation robotics companies can attract capital when their technology speaks for itself.

Spirit AI's Landmark Funding Milestone in China's AI Landscape

The past several months have seen China's startup ecosystem break new ground in robotics and AI investment, but Spirit AI's back-to-back fundraise represents a particularly striking development in this ongoing story. Founded as recently as February 2024, the company has compressed what would normally take years of capital-raising into an aggressive timeline, reflecting both the urgency of the embodied intelligence market and the sheer strength of its early-stage execution. The two consecutive funding rounds were announced on February 24, 2026, and together represent a combined total close to the 2 billion yuan mark — a figure that has sent strong signals across the global tech investment community and reaffirmed China's growing ambition in the embodied AI space. The timing could not be more strategic, as nations around the world race to establish dominance in the intersection of physical robotics and large-scale AI models.

What makes this instance of AI funding especially noteworthy is not merely the quantum of capital raised, but the extraordinary diversity and credibility of the investor base assembled in such a compressed timeline. The company has managed to bring together financial capital, strategic industrial investors, and state-backed funds under a single investment umbrella — a configuration that many startups spend years trying to engineer. The freshly raised capital will be directed primarily toward scaling Spirit AI's foundational embodied AI models, expanding real-world data pipelines, and deepening operational collaboration across its growing network of industrial partners. In an era where AI funding news is predominantly dominated by pure-software companies building large language models and generative tools, Spirit AI's hardware-plus-model approach has clearly captured the imagination of investors who are looking for companies genuinely capable of bridging the digital and physical worlds in a meaningful and commercially viable way.

The announcement arrives at a moment when the global race for dominance in humanoid robotics and embodied intelligence has intensified dramatically. Companies from the United States, Europe, and across Asia are all competing to develop robots that can function autonomously in complex, real-world environments without relying on pre-programmed instructions for every scenario. China's central government has also made embodied intelligence a national strategic priority, channeling substantial policy support and state funding toward companies developing next-generation robotic systems. Spirit AI's fundraise, therefore, is not just a corporate milestone — it is a reflection of a much larger macro-level conviction that embodied AI will become central to the next phase of global industrial transformation, and that China intends to be at the forefront of that shift.

A Power-Packed Investor Consortium: Who's Backing Spirit AI?

The roster of investors that participated in Spirit AI's latest funding rounds reads like a who's who of China's most influential capital networks. Among the newly committed participants, Yunfeng Capital — a leading state-affiliated investment institution with deep ties to China's financial establishment — took part, signaling that this is not merely a private market bet but one that carries the implicit endorsement of the country's broader economic apparatus. Chaos Investment, associated with prominent investor Ge Weidong, and HongShan, one of China's most respected and well-established venture capital firms, also joined the round, adding further prestige and strategic weight to an already formidable cap table. The presence of these names alone would make this AI funding event newsworthy, but the story becomes even more compelling when you look at the full investor picture.

On the industrial investor side, Synstellation Capital and TCL Capital participated in the rounds, reflecting the strategic interest that China's established manufacturing and consumer electronics sectors have in owning a stake in the future of humanoid robotics. State-owned capital vehicles from Chongqing and Hangzhou — two cities that have been particularly aggressive in backing emerging deep technology companies — also committed funds, reinforcing the view that Spirit AI's mission has achieved broad government buy-in at both central and regional levels. Additionally, 360 Capital and HouXue Capital rounded out the new investor lineup, bringing cybersecurity and consumer technology perspectives into a syndicate that is already rich with industrial domain expertise. These names collectively represent a significant concentration of financial firepower, sectoral knowledge, and institutional credibility that Spirit AI can draw on as it scales its operations in the years ahead.

Existing investors were equally enthusiastic about doubling down on their positions. ShunWei Capital, Prosperity7, and Fortune Capital all chose to recommit fresh capital, a signal of sustained and growing confidence in Spirit AI's trajectory since its founding. Perhaps the most strategically compelling dimension of this AI funding story, however, is the composition of Spirit AI's broader shareholder ecosystem, built up across all of its funding rounds to date. The company has assembled a strategic investor base that includes CATL, the world's largest battery manufacturer; TCL, a globally recognized electronics and appliance giant; JD.com, China's second-largest e-commerce platform and one of its most sophisticated logistics operators; China Merchants Innovation Investment Management; and consumer technology powerhouses Huawei and Xiaomi. This kind of full-spectrum industrial ecosystem — where suppliers, manufacturers, retailers, distributors, and technology platform companies all sit on the same cap table — is extraordinarily rare and gives Spirit AI access to real-world deployment environments, proprietary operational data streams, and go-to-market channels that most robotics startups can only dream of. For anyone following AI funding news in the deep tech space, the construction of this ecosystem-level investor base is arguably the single most strategically significant dimension of Spirit AI's story.

Inside the Technology: Moz1 Humanoid Robot and the VLA Model

Spirit AI's funding success is ultimately grounded in a technology foundation that has already demonstrated tangible, measurable results in real-world environments. At the heart of its product lineup is the Moz1, a full-body force-controlled humanoid robot built on the company's proprietary Vision-Language-Action (VLA) model — a class of AI architecture that integrates visual perception, language understanding, and physical action planning into a single unified system. The Moz1 features 26 degrees of freedom, a technical specification that allows it to replicate a wide range of human-like physical movements with considerable precision and fluidity. The robot also incorporates high power-density force-control joints and high-speed whole-body control algorithms, enabling it to achieve a payload-to-weight ratio of 1:1, which means the robot can carry loads equivalent to its own body weight. In practical terms, this dramatically expands the range of industrial tasks the robot can perform and makes it suitable for demanding physical environments where prior generations of robotic systems would have been inadequate.

The VLA model that powers Moz1 is not simply a motion controller for the hardware — it is a generalized intelligence layer that allows the robot to reason about its environment, interpret instructions, and make autonomous decisions in real time. This is the fundamental distinction between Spirit AI's approach and the earlier generations of industrial automation, which relied entirely on pre-programmed, hard-coded instructions for every specific task a robot was expected to perform. Spirit AI's system, by contrast, is designed to generalize across novel and unstructured situations without requiring task-specific reprogramming, which is what makes it genuinely useful and cost-effective in the kind of dynamic, unpredictable environments that characterize real factories, warehouses, and retail stores. The implications of this distinction are enormous — it means that a single model and hardware platform can, in principle, be deployed across a wide variety of use cases without the custom engineering work that has historically made robotics deployments so expensive.

Earlier in 2026, Spirit AI made a significant strategic move by open-sourcing Spirit v1.5, its latest generation foundational model. The release drew widespread attention within the global AI and robotics research community because of the model's impressive zero-shot generalization capabilities. In technical terms, zero-shot performance refers to a model's ability to successfully carry out tasks it has never explicitly been trained on, drawing instead on broader patterns and physical intuitions learned during large-scale pretraining. In the case of Spirit v1.5, this means the model can tackle complex physical manipulation challenges — such as wiping surfaces, operating hinged doors and containers, or handling flexible and deformable objects — without any task-specific fine-tuning or additional training data collection. This is a substantial engineering achievement and one that places Spirit AI in a rare class of embodied AI developers globally. The company attributes this capability to what it calls its "data pyramid" methodology, which prioritizes large-scale pretraining on human internet video as a foundational knowledge base. Rather than relying exclusively on proprietary robot-specific training data — which is expensive and time-consuming to collect at scale — Spirit AI leverages the enormous wealth of information embedded in naturally occurring videos of humans performing everyday tasks. This approach allows its models to develop an intuitive understanding of how physical objects behave and how intelligent agents navigate real-world spaces, achieving strong generalization performance with fewer model parameters and at significantly lower computational cost than competing approaches.

Real-World Deployment: From CATL's Factory Floor to JD.com's Retail Shelves

One of the most compelling and differentiating aspects of Spirit AI's story is that the company is not merely a research-stage startup working toward theoretical milestones — it has already transitioned into real-world commercial deployment, and the outcomes have been genuinely remarkable. The company's humanoid robot "Xiaomo" has been deployed at CATL's Zhongzhou production base, one of China's most advanced and demanding battery manufacturing facilities. At this site, Xiaomo has been integrated as a core piece of equipment on the battery assembly line, where it has contributed to the zero-defect production of close to 1,000 battery units. What makes this deployment truly significant is that the robot has reportedly matched — and in some instances surpassed — both the pace and precision of skilled human workers on the assembly line, a benchmark that the industrial robotics industry has long considered a critical and elusive threshold for commercial viability. This is not a controlled lab demonstration or a pilot with minimal volume — it is a fully operational deployment in a high-stakes manufacturing environment with real quality and throughput requirements.

The deployment at CATL is significant for multiple interconnected reasons that go beyond the immediate technical validation it provides. First, it proves that Spirit AI's technology stack is mature enough for continuous, mission-critical industrial operations rather than the curated, low-stakes demonstrations that many robotics companies rely on to attract investor attention. Second, CATL's Zhongzhou production base is one of the most demanding manufacturing environments in China, meaning that success there effectively certifies Spirit AI's technology for a wide range of adjacent industrial applications. Third, and perhaps most important from a business development perspective, the CATL deployment provides Spirit AI with a high-value, real-world reference case that can be leveraged to win new customers across other manufacturing sectors where automation and labor efficiency are major priorities. In the context of AI funding news from China's robotics sector, this kind of production-grade commercial deployment makes Spirit AI's investment case uniquely credible compared to the many competitors that remain in pre-commercial development phases.

Beyond the factory floor, Spirit AI has also established an exciting commercial partnership with JD.com to deploy its Moz robots in consumer-facing retail settings. In this context, the robots are engaged in customer interaction and product demonstration tasks — roles that demand an entirely different kind of intelligence profile than factory automation. Retail environments require natural and intuitive communication, real-time spatial awareness in crowded and unpredictable spaces, and the ability to adapt continuously to the wide variety of human behaviors encountered in a live store setting. Building on the success of this initial deployment, Spirit AI and JD.com are now jointly exploring how JD Cloud infrastructure and the Joyinside large model can be scaled across large-scale retail networks, creating a tighter integration between model development and commercial deployment at scale. This kind of vertical integration strategy — from foundational model research all the way through to deployment across both industrial and consumer-facing environments — is what fundamentally distinguishes Spirit AI's commercialization approach from many of its peers in the global robotics and embodied AI industry.

The Road Ahead: Embodied Intelligence and China's Global AI Ambitions

The pace of AI funding news flowing out of China's robotics sector in early 2026 reflects a broader and accelerating transformation underway in how the country is approaching its technological future on the global stage. Embodied intelligence — the discipline that seeks to give artificial intelligence a physical form capable of operating effectively and autonomously in the real world — has emerged as one of the most competitive and strategically consequential frontiers in global technology. Governments, sovereign wealth funds, and private investors around the world are pouring capital into companies that can close the gap between digital AI models and physical robotic systems capable of performing meaningful work across industries. Spirit AI, with its deep industrial partnerships, a production-proven hardware platform, an open-source model ecosystem, and documented commercial deployments, is now positioned as one of the most credible and well-resourced contenders in this race.

With the fresh capital from its latest AI funding rounds in hand, Spirit AI has outlined clear and ambitious priorities for the road ahead. The company plans to accelerate the development of its next-generation foundational embodied AI models, with a particular emphasis on expanding the scale and diversity of its real-world training data. This will involve building out deeper data collection partnerships with its existing industrial collaborators — including CATL, JD.com, and others across its shareholder ecosystem — to create powerful feedback loops between real-world deployment experience and continuous model improvement. The company also intends to deepen its collaboration with its manufacturing and retail partners already embedded in its ecosystem, while potentially expanding into new verticals such as logistics, healthcare, hospitality, and elder care as the capabilities of its systems continue to mature and the cost of deployment continues to decrease with scale.

It is worth putting Spirit AI's funding achievement into a broader global context to fully appreciate its significance. The total quantum of nearly 2 billion yuan raised across consecutive rounds places it among the best-funded embodied AI startups not just in China, but by any measure globally. The presence of both high-profile private capital and state-backed institutional investors in its shareholder base means that the company has a level of financial stability and institutional support that many of its rivals simply cannot match — a critical advantage in a capital-intensive field where hardware development timelines are inherently long and revenue ramp-up can take several years to fully materialize. The fact that all of this has been accomplished by a company that was only founded in February 2024 speaks to an extraordinary pace of execution, a clear and compelling long-term vision, and a founding team that has managed to articulate that vision in a way that has resonated widely and deeply across China's investment community. The embodied intelligence landscape is moving fast, and Spirit AI, with its rare and hard-to-replicate combination of technical depth, industrial ecosystem, and investor backing, appears exceptionally well-positioned to help define where this transformative industry goes next.