Leju Robotics

Leju Robotics closed a round of approximately $200 million in October 2025. Investors in this round included Shenzhen Investment Holdings Capital, Shenzhen Longhua Capital, Qianhai Foundation Investment, Oriental Precision, and CITIC Jingshi.

The company previously raised a $350 million Series B round with participation from Tencent, Hongtai Fund, and Shenzhen Capital Group. Earlier funding included a $7 million strategic investment from Tencent, plus seed and Series A rounds in 2016 and 2017 respectively.

Leju builds three classes of bipedal robots that share in-house torque motors and balance algorithms.

KUAVO is the full-size humanoid standing 170 cm tall and designed for real-world applications. It walks at 4.6 km/h on sand, grass, and factory floors, can jump 20 cm, and has 360 Nm joints for lifting moderate payloads. The robot runs on Huawei's HarmonyOS with the Pangu AI model and supports voice commands, task planning, and remote programming via HDMI, USB, or cloud interfaces.

AELOS robots are 30-40 cm desktop units with 17-22 degrees of freedom for K-12 education, universities, and RoboCup competitions. They connect via Bluetooth and Wi-Fi, support programming languages from Scratch to C++, and include IMU sensors, depth cameras, and modular grippers. Students can drag-and-drop actions through a choreography app or write custom firmware.

PANDO is a palm-sized panda-shaped coding toy that rolls, bows, and responds to voice commands. Children start with Scratch programming, advance to Python, and can load lesson plans from phones or PCs. IR sensors support gesture following and obstacle avoidance with 2-3 hours of battery life per session.

Research labs use KUAVO with ROS 2 integration and an open CAN bus architecture to mount custom end-effectors and capture gait data. Factories deploy the robots as bilingual tour guides or for material handling, loading workflow templates through cloud consoles. Smart home pilots connect the robots to IoT devices for voice-controlled automation tasks.

Leju runs a B2B model selling hardware units with embedded software and ongoing service contracts. The company vertically integrates critical components, manufacturing its own 360 Nm torque servos to control costs and differentiate from competitors that rely on third-party actuators.

Revenue comes from direct unit sales across three price tiers. Full-size KUAVO humanoids generate the highest margins despite recent price compression, while AELOS education kits provide recurring revenue through school refresh cycles. PANDO units serve as entry-level products for younger students entering the product line.

The go-to-market approach varies by segment. Education sales flow through established distributor networks built over years of serving schools with robotics curricula. Industrial humanoid sales involve direct enterprise relationships, pilot programs, and partnerships with systems integrators for deployment in factories, showrooms, and smart buildings.

Leju's integration with Huawei's HarmonyOS gives access to Pangu AI capabilities without the licensing costs of competing platforms. The arrangement shortens time to deploy voice control, computer vision, and task planning features that would otherwise require extensive in-house AI development.

Deployed robots generate usage data to improve AI models, and the education segment builds long-term relationships with institutions that may later adopt industrial humanoids as costs decline.

Tesla and Agility Robotics exemplify the manufacturing-focused approach to humanoid robotics. Tesla leverages its automotive production expertise and large-scale data collection through its vehicle fleet to train AI models, though the Optimus robot remains in early development stages. Agility Robotics focuses on pilot programs with large corporations, with strategic partnerships and data-driven AI development for warehouse and logistics applications.

Both companies prioritize control of the full manufacturing stack and real-world deployment data collection, similar to Leju's vertical integration strategy with torque motors and HarmonyOS integration.

1X Technologies and Sanctuary AI prioritize AI development over hardware manufacturing, using real-world data to improve intelligent capabilities while pursuing scalable manufacturing partnerships. These companies focus on foundation model training and end-to-end AI adaptability in dynamic environments.

This approach contrasts with Leju's hardware-first strategy, though both treat real-world data collection as key to improving robot performance and expanding addressable use cases.

Fourier Intelligence and Unitree Robotics pursue ultra-low-cost strategies backed by state funding and aggressive pricing. Unitree's new R1 model priced at approximately $5,500 sets new entry-level expectations for hobbyist and educational users, creating pricing pressure on premium educational robots like Leju's AELOS family.

Fourier Intelligence combines rehabilitation robots with open-source humanoid development, targeting universities and hospitals with complete development kits. Both companies benefit from Chinese manufacturing scale and government support, which pressures higher-cost competitors to differentiate through software capabilities and service quality rather than hardware pricing alone.

Leju's product portfolio already spans multiple robot form factors and price points, creating upsell opportunities within existing customer relationships. The company's joint venture with Haichen to develop the Sailian Aura industrial-grade humanoid targets factory logistics, an adjacent market worth over $30 billion where mobile manipulators are replacing fixed conveyor systems.

The 5G-enabled demonstration robot developed with China Mobile and Huawei supports ultra-low-latency teleoperation and cloud inferencing capabilities. This enables robot-as-a-service business models for malls, airports, and smart city operators who prefer operational expenses over capital equipment purchases.

Management's roadmap includes an open humanoid robot platform that would generate recurring software revenue from third-party developers, shifting the revenue mix from hardware manufacturing toward software and ecosystem coordination.

Leju's cost reduction roadmap targets a price point of approximately $20,000 by late 2025, opening security patrol, delivery, and consumer segments measured in millions of potential units rather than thousands. This shifts focus from B2B research and industrial applications toward mass consumer markets.

The partnership with China Telecom packages robots for elder care, safety monitoring, and family companionship, expanding the total addressable market from research laboratories toward consumer IoT applications. Early pilots in smart apartments include integration with home automation systems and voice-controlled assistance.

The national joint laboratory with Peking University aims to make Leju's platform a common teaching and research platform, capturing long-tail university and vocational training budgets while building relationships with the next generation of robotics researchers and engineers.

International exhibitions including Mobile World Congress in Barcelona provide entry points into European and North American markets where labor shortages create acute demand for general-purpose automation. These markets typically support higher price points than domestic Chinese sales, potentially improving unit economics for export models.

The pre-IPO financing specifically earmarks funds for deepening industry chain layout and export certification, reducing capital constraints on international scaling. Leju's integration with globally available platforms like HarmonyOS and standard robotics frameworks like ROS 2 facilitates international deployment compared to China-specific software stacks.

Price compression: Aggressive pricing from competitors such as Unitree, which launched a humanoid robot at $5,500, adds downward pressure on Leju's education robot margins and accelerates required cost reduction in industrial models. Leju faces tradeoffs between quality and features versus price, especially against state-subsidized rivals.

Technology commoditization: As AI models standardize and hardware components become widely available, differentiation from vertical integration may erode. The shift toward foundation models and open-source robotics software lowers barriers to entry for new competitors and could commoditize the software stack that underpins Leju's products.

Market timing: The humanoid robotics industry remains in early stages with uncertain adoption timelines across target markets. If enterprise customers delay humanoid deployments or consumer acceptance develops more slowly than projected, Leju's growth trajectory and valuation could face pressure, particularly given the capital-intensive nature of robotics hardware development.