Kinetix AI Secures $75 Million Angel Funding to Accelerate Its AI‑Hardware Feedback Loop

Closing the Gap Between Digital Intelligence and Physical Action

Engineers have long observed that AI systems capable of generating text instantly still fail to handle delicate objects without damage.

The core problem, according to the source, is the missing continuous feedback loop that aligns software cognition with mechanical execution.

Kinetix AI positions its solution as a seamless loop that couples multimodal data capture, embodied foundation models, and high‑fidelity hardware.

This loop is described as an “intelligence flywheel” where each physical interaction refines the AI model, which in turn upgrades the robot’s capabilities.

By eliminating the need to redesign hardware after software updates, the approach promises faster iteration and more reliable performance.

Funding Surge and Leadership Team

Kinetix AI announced an angel round of over 500 million yuan, roughly $75 million, to fund its closed‑loop ecosystem.

The round was led by Vertex Ventures, the venture arm of Singapore’s Temasek Holdings.

F&G Venture and Wanshi Capital also participated, contributing to the record‑breaking financing.

Founded in September 2025, Kinetix AI has grown to nearly 200 employees in less than two years.

CEO Yu Jie, who previously commercialized Huawei’s autonomous mining trucks, heads the company.

Associate dean Luo Ping from Hong Kong University brings academic depth to the venture.

Co‑founder Zheng Cunyuan, formerly robotics head at XPeng, adds industry‑scale robotics experience.

Premium Human‑Like Robots and Real‑World Validation

Kinetix’s flagship robot, KAIBot, stands 1.73 meters tall and offers 115 degrees of freedom, matching human articulation.

The company emphasizes that the robot’s premium design, though costly and supply‑chain intensive, is essential for true embodied intelligence.

Co‑founder Zheng Cunyuan argues that “starting with cheap, subpar hardware is just foolish,” underscoring the strategic focus on quality.

By mirroring human form factor, KAIBot can directly map to human motion data, allowing AI models to transfer without loss.

The hardware suite also includes the KAI Hand, a highly dexterous manipulator designed for fine‑grained tasks.

Kinetix demonstrated the system’s dynamism at the 2026 World Humanoid Robot Games, where its robot played table tennis against world champion Ding Ning.

This public showcase highlighted the robot’s ability to execute complex, real‑world motions using the closed‑loop feedback architecture.

The company’s long‑term vision is to create a premium category of robots ready for complex deployment in everyday environments.

Why This Matters: Kinetix AI’s $75 million angel round equips it to build a continuous AI‑hardware feedback loop that could close the performance gap between digital intelligence and physical manipulation.