WUJI HAND

Source-backed capabilities

• Operating environments: Physical research setup
• Control modes: Learned Policy, Learned Policy With Operator Motion Commands, Dedicated Learned Recovery Policy, Learned Pushing Policy With Visual Feedback, Hug Grasp Prediction And Open Loop Execution
• Fingertip crawling

  Research demonstration - 2026-09-15 - Crawl untethered across 14 indoor and outdoor surfaces using an onboard learned policy. - environment: Physical research setup

  Source
• Fall recovery

  Research demonstration - 2026-09-15 - Recover onto the fingertips after a fall using a separate recovery policy. - environment: Physical research setup

  Source
• Object manipulation

  Demonstrated - 2026-05-06 - Human-scale proportions and back-drivability support transfer from human demonstrations to robotic manipulation.

  Source
• Cube reorientation

  Research demonstration - 2026-09-27 - Rotate a tagged cube within the hand using a learned policy and camera feedback. - environment: Physical research setup

  Source
• Self-supported cube pushing

  Research demonstration - 2026-09-15 - Approach a cube and push it to a target using overhead visual feedback. - environment: Physical research setup

  Source
• Household grasping

  Research demonstration - 2026-06-15 - Grasp household objects on a mobile manipulator with HUG; the research setup succeeds in 62% of 300 trials. - environment: Physical research setup

  Source

Exact-product configurations are shown without inferring supply, purchase, customer status, standard fit, deployment or partnership.

• Delta-0 whole-body manipulation demonstration

  2026-09-30 - Physical Demo - Connected - Physical - Delta Intelligence demonstrated Delta-0 using WUJI HAND to steady and accurately place a record while coordinating the hands, arms and torso.

  Source
• Modified YOR mobile manipulator with AgileX NERO arm

  Research Experiment - Mounted - Physical - HUG uses the original Wuji hand on a modified YOR mobile manipulator with an AgileX NERO arm for household grasping. The mobile evaluation spans 300 trials.

  Source

• HandsOn: A Large-Scale Simulation Benchmark for Generalizable Dexterous Manipulation

  2026-10-19 - Research Publication - HandsOn provides 66 dexterous-manipulation simulation tasks with 18,000 demonstrations. Its quantitative evaluations use a ROKAE AR5-08 and first-generation Wuji hands, with qualitative transfer to physical hardware.

  Source
• Gen2Dex: Learning Dexterous Manipulation from Simulation-Initialized Generated Video Demonstrations

  2026-10-17 - Research Publication - Gen 2Dex turns generated manipulation videos into executable dexterous trajectories using known simulated geometry, hand-object tracking and residual reinforcement learning. A backward curriculum corrects imperfect motion references. The university team evaluates simulated hands and transfers optimized trajectories to a physical Wuji Hand mounted on an xArm.

  Source
• Wuji-MJLab: Cube Reorientation and Pen Spinning

  2026-09-27 - Research Publication - Cube reorientation supports both Wuji hand generations with separate hardware and checkpoint configurations. Reference-tracking pen spinning targets Wuji Hand 2 and transfers learned policies to the physical hand. The asset bundle includes deployment checkpoints, hand jigs, cube and pen geometry, and 40 motion clips.

  Source
• Fingers as Legs: Learning Self-Supported Locomotion and Manipulation with an Anthropomorphic Hand

  2026-09-15 - Research Publication - ETH Zurich researchers use an off-the-shelf WUJI HAND for untethered crawling, steering, fall recovery, keyboard input and object pushing.

  Source
• CoorDex: Coordinating Body and Hand Priors for Continuous Dexterous Humanoid Loco-Manipulation

  2026-06-22 - Research Publication - CoorDex coordinates body and hand priors for simulated continuous humanoid loco-manipulation. Quantitative task evaluation uses G1 with WUJI hands in Isaac Lab; separate physical G1 with Dex3-1 visualizations replay recorded joint trajectories.

  Source
• TopoRetarget: Interaction-Preserving Retargeting for Dexterous Manipulation

  2026-06-15 - Research Publication - The authors introduce TopoRetarget, an interaction-preserving retargeting framework that uses a single set of parameters across diverse retargeting conditions while maintaining task-relevant hand-object interaction and adapting human demonstrations to dexterous robot hands.

  Source
• Human Universal Grasping

  2026-06-15 - Research Publication - Introduces HUG, a flow-matching model that predicts diverse human grasps from a single RGB-D image.

  Source
• Kilohertz-Safe: A Scalable Framework for Constrained Dexterous Retargeting

  2026-03-31 - Research Publication - This paper proposes a scalable motion retargeting framework that reformulates the nonlinear retargeting problem into a convex quadratic program in joint differential space.

  Source

• Wuji expands Japanese distribution through HumansX

  2026-10-06 - Partnership - Japanese-language setup and technical support cover both the dexterous hand and tactile data glove. The offering includes contracted hand-motion and tactile-data collection for research and AI training. Integration guidance covers SDKs, ROS 2, simulation, teleoperation and motion retargeting.

  Source
• Genesis AI demonstrates GENE-26.5 with custom Genesis Hand 1.0

  2026-05-06 - Demonstration - The demonstration used Genesis Hand 1.0, a custom design rather than the catalog Wuji Hand. Wuji's corrected account identifies its role as a manufacturing collaborator. Human-scale proportions and back-drivability support transfer from human demonstrations to robotic manipulation.

  Source
• Wuji publishes technical documentation for its first-generation robotic hand

  2025-09-14 - Market Signal - The official FAQ documents 20 total degrees of freedom, including 15 active and five passive degrees of freedom. Wuji describes an electric direct-drive architecture with integrated motor, drive and reducer modules. The documentation covers communication, control, sensing and integration information for developers.

  Source

Market context

High-DoF dexterous hand with public weight, force, load, drive and interface specifications.

• Target environment: Humanoid robots, embodied-AI labs and research manipulation rigs
• Workflow context: Dexterous grasping, teleoperation, robot-hand integration and manipulation benchmarking
• Customer context: Teams evaluating compact hands for humanoid or arm-mounted manipulation.
• Deployment model: Supplier product page and contact/sales path.
• Commercial maturity: Commercial-facing product page; buyer terms are not publicly detailed.
• Adoption constraints: Integration depends on USB/control access, durability, mechanical fit and validation under repeated manipulation cycles.
• Adjacent products: Unitree Dex5-1, DexH13, AIDIN-Hand, RealHand O6, SCHUNK SVH

Comparable Systems

• Aero Hand Open

  Chestnut Robotics - Dexterous Hand

• OmniHand O10 2025

  AgiBot - Hand

• OmniHand Pro 2025

  AgiBot - Hand

• AIDIN-Hand

  Aidin Robotics - Robotic Hand

• Amazing Hand

  Pollen Robotics - Robotic Hand

• Apex Hand

  Dexcel Robotics - Dexterous Hand