X Square Robot

Summary

About

About X Square Robot

X Square Robot links robot-learning software with hardware and tools for collecting manipulation data. Quanta X1 Pro combines a wheeled base, two arms and a vertically adjustable workspace so researchers can work from floor level to tables and shelves. WALL-A is integrated with the platform alongside native ROS 2 support.

Operators can demonstrate tasks through matching robot arms or VR controls, then capture and unpack the resulting data through a graphical interface. Python SDKs and other APIs support custom applications. This connects demonstration collection, model development and execution on the same research platform; the wider hardware portfolio includes Quanta X2 and the ArtiXon dexterous hand.

PRICE
Range 1Y
Return
High
Low
Updated

Articles

Products

Deployments

  • Customer site

    2026-09-16 - X Square Robot's QUANTA X1 Pro is operating on the logistics line at Lululemon's Wuhan Distribution Center, handling and organizing parcels as part of daily warehouse operations. - Wuhan, China

    Source

Supply

Supply chain relationships

  • D-Robotics

    Inbound - Compute & Control - X Square Robot adopted D-Robotics Sunrise S600 for embodied-AI computing

    Source

Funding

Investors

  • Meituan Strategic Investment - Lead - Corporate / CVC - 2 funding events - repeat backer - strategic/corporate - first seen 2025-05-12 - latest 2025-09-08
  • Xiaomi Strategic Investment - Lead - Unknown - 2 funding events - repeat backer - first seen 2026-04-21 - latest 2026-06-29
  • Alibaba Cloud - Lead - Corporate / CVC - 1 funding event - strategic/corporate - first seen 2025-09-08
  • Beijing Robot Industry Fund - Participant - VC - 1 funding event - first seen 2025-02-17
  • ByteDance - Participant - Unknown - 1 funding event - first seen 2026-01-13
  • CAS Investment - Participant - Angel / Individual - 1 funding event - first seen 2025-09-08
  • China Development Bank Capital - Participant - VC - 1 funding event - first seen 2025-09-08
  • China Growth Capital - Participant - VC - 1 funding event - first seen 2025-05-12
  • China Mobile - Participant - Angel / Individual - 1 funding event - first seen 2026-06-29
  • CICC Capital - Participant - VC - 1 funding event - first seen 2026-06-29
  • Dachen Caizhi - Participant - Angel / Individual - 1 funding event - first seen 2026-06-29
  • GF Xinde - Participant - Angel / Individual - 1 funding event - first seen 2025-05-12
  • HongShan - Lead - Unknown - 1 funding event - first seen 2026-04-21
  • IDG Capital - Participant - VC - 1 funding event - first seen 2026-06-29
  • Legend Capital - Participant - VC - 1 funding event - first seen 2025-02-17
  • Lenovo - Participant - Unknown - 1 funding event - first seen 2025-09-08
  • Lightspeed China Partners - Participant - VC - 1 funding event - first seen 2025-02-17
  • Meituan Longzhu - Participant - Corporate / CVC - 1 funding event - strategic/corporate - first seen 2025-05-12
  • National AI Industry Investment Fund - Participant - VC - 1 funding event - first seen 2026-06-29
  • SAIC Financial Services - Lead - Unknown - 1 funding event - first seen 2026-02-28
  • Sequoia China - Participant - Angel / Individual - 1 funding event - first seen 2026-06-29
  • Shenqi Capital - Participant - VC - 1 funding event - first seen 2025-02-17
  • Source Code Capital - Participant - VC - 1 funding event - first seen 2026-06-29
  • Yunqi Partners - Participant - VC - 1 funding event - first seen 2025-05-12

Team

  • Wang Hao - Co-founder
  • Wang Qian - Founder and Chief Executive Officer - Founder-CEO leading X Square Robot through embodied-AI model releases, robot-platform development and private financing rounds.

Research

  • X Square open-sources X2Real robot-policy evaluation runtime

    2026-10-10 - Research Publication - The technical report defines 44 hierarchical tasks across 10 capability dimensions and reports 0.84 correlation between simulation and real-robot task-progress scores. The benchmark separates training and evaluation settings and varies appearance, task configuration and robot embodiment. ManaEnv manages simulation and scoring while separate Policy Space services run policy inference.

    Source
  • X Square introduces WALL-SS long-horizon robotic world model

    2026-08-26 - Research Publication - WALL-SS generated action-conditioned streaming rollouts up to 60 seconds while retaining recent history at fine resolution and compressing older state under a bounded memory budget. Across 600 matched simulated and physical rollouts on six manipulation tasks, task-level success rates had mean absolute error of 0.062 and Pearson correlation of 0.93. A co-trained action expert controlled a physical bimanual platform across seven tabletop tasks and reached average Task Progress of 69.1, ahead of the three reported comparison policies.

    Source
  • X Square Robot open-sources HOST one-shot skill-acquisition framework

    2026-08-03 - Research Publication - HOST was evaluated across 50 novel manipulation tasks and achieved 62% average success on the eight-task baseline-comparison set. Average skill-acquisition time was 29 seconds, with a 43-point advantage over the strongest compared method without parameter updates. The project page, implementation code and paper are publicly available; the work was developed with researchers from Tsinghua University and Beijing Institute of Technology.

    Source
  • XRZero-G0 robot-learning dataset release

    2026-06-10 - Research Publication - XRZero-G0 was published as an official X Square Robot release. The dataset adds a separate training-data milestone beyond the company's WALL model releases. The research strengthens XRZero-G0 robot-learning dataset release's technical work in commercial automation and autonomous-system workflows.

    Source
  • WALL-WM world-model release

    2026-05-29 - Research Publication - WALL-WM was published as an official X Square Robot release. The release shifted the company's robot world-model work toward event-level modeling. The event adds a distinct world-model milestone after WALL-OSS-0.5.

    Source
  • X Square Robot open-sources Wall-OSS-0.5 VLA model

    2026-05-26 - Research Publication - Wall-OSS-0.5 explores whether pretrained vision-language-action models can produce robot capabilities directly testable on physical hardware before task-specific fine-tuning. Technical highlights include gradient-bridged co-training, a Vision-Aligned RVQ Action Tokenizer, action-space supervision and the DMuon distributed optimizer. In zero-shot real-robot evaluation, the pretrained checkpoint achieved task-progress scores above 80 on Block Sorting, Fruit Sorting, Ring Stacking and Rope Tightening.

    Source

Timeline

Events

Relations

Relationships

  • Lululemon

    Warehouse Deployment Customer - X Square's QUANTA X1 Pro handles and organizes parcels on Lululemon's Wuhan distribution-centre logistics line.

  • 58.com

    Home Cleaning Service Partner - X Square Robot and 58.com launched a Shenzhen home-cleaning service pairing cleaners with X Square robots.

Context

Comparable profiles

  • Physical Intelligence

    comparable profile - Vla Robot Policy Developer

  • LimX Dynamics

    comparable profile - Biped Robot Development Platform

  • MiraiBit

    comparable profile - Vla Robot Policy Developer

  • MOTOMIND

    comparable profile - Vertically Integrated Humanoid Builder

Shared workflows

Context tags

  • Generalist robot policy - Capability - WALL-OSS model release exposes generalist robot-policy work.
  • VLA robot policy developer - Peer Group - X Square Robot publishes WALL VLA and embodied foundation models for real-robot action and manipulation: https://x2robot.com/en/research
  • Whole-body humanoid control - Capability - Humanoid model and hardware context supports whole-body robot-control relevance.
  • Vertically integrated humanoid builder - Peer Group - Company develops both robot hardware and foundation-model/control stack.
  • Whole-stack humanoid development - Workflow - Profile combines robot hardware, model training and deployment work.
  • Humanoid builders - Stack Layer - Company builds humanoid robot systems.
  • China robot builder - Peer Group - China-headquartered robot builder.
  • Biped robot development platform - Peer Group - Robot platforms support embodied-intelligence development.