
MindOne Robotics (MindOn Everything) is a Shenzhen-based Chinese startup building a universal robot intelligence foundation model. It trains AI on human-centric data and deploys it across heterogeneous robot platforms, including the Unitree G1.
MindOne Robotics, also operating as MindOn Everything, is a Chinese robotics startup based in Shenzhen, founded in May 2025 by Qingxu Zhu. The company does not build hardware — it builds a universal robot foundation model (Mind-0), hardware-agnostic by design, intended to drive both humanoids (e.g., Unitree G1) and stationary dual-arm systems with a single AI model.
MindOne's key technical bet is moving away from robot teleoperation as a source of training data. Instead, the company trains its models on human-centric data captured via whole-body motion capture, handheld devices, and egocentric cameras. The architecture is split into two layers — a high-level model handling scene understanding, task reasoning and behavior generation, and a low-level Whole-Body Action Model controller translating intentions into physical motion specific to each robot embodiment. A lightweight Real-World Execution Compensation Model bridges the sim-to-real gap and, according to the company, achieves sub-1 cm manipulation accuracy on the Unitree G1 — a platform typically known for limited arm precision.
MindOne drew industry attention with two demonstrations: a November 2025 viral video of a Unitree G1 autonomously performing complex household chores (no speed-ups, no teleoperation), and a June 2026 heterogeneous-fleet demo coordinating two Unitree G1 humanoids and two stationary dual-arm rigs through a full logistics workflow (shelf picking, packing, tape sealing) — all driven by a single AI model. The company is hiring and seeking deployment partners in logistics, warehouse automation and home robotics.
Founders
Founder of MindOne Robotics, the public face of the company's demonstrations (including the November 2025 viral video of a Unitree G1 performing household chores and the June 2026 coordinated heterogeneous robot fleet demo).
Classification
External links