
ORCA Dexterity
PrivateORCA Dexterity from ETH Zurich develops open source robotic hands with integrated tactile sensing.
Dexterous Hands
About ORCA Dexterity
ORCA Dexterity is an open-source robotics project from ETH Zurich developing affordable, high-performance robotic hands for research and embodied AI. The project combines dexterous manipulation, integrated tactile sensing, and open hardware to make advanced robotic hands more accessible to researchers, developers, and robotics laboratories. Its robotic hands are designed to be largely 3D printable while incorporating integrated sensing and modular components for manipulation research. With multiple models ranging from approximately $1,500 to $6,100, ORCA Dexterity offers a cost-effective platform for studying grasping, tactile perception, and human-like manipulation in humanoid robots and other robotic systems. Tagline
Technical Profile
Platform maturity, autonomy stack, and flagship-system specifications in one view.
Technology Snapshot
A concise view of platform maturity and deployment footprint.
AI Stack
Capabilities and model layers highlighted by the company.
ORCA Hand Specs
Published operating specifications for the lead system.
Latest Updates
In March 2026, released three models: orcahand lite ($1,500), standard ($3,500), and touch with 83 taxels per fingertip ($6,100).
ORCA Dexterity Team
ORCA Dexterity History
- 2024
ORCA Dexterity emerges from ETHs Soft Robotics Lab
Emerged from ETH Zurich's Soft Robotics Lab led by Robert Katzschmann. Designed a 17-DOF tendon-driven hand assemblable in under 8 hours for ~$2,000.
ETH ZurichSoft Robotics Lab - April 2025
Open-source publication formalizes the platform
Published as academic paper in April 2025. All CAD files and code released under permissive open-source licenses.
Open sourceAcademic paper - March 2026
The hand lineup expands into multiple commercial tiers
In March 2026, released three models: orcahand lite ($1,500), standard ($3,500), and touch with 83 taxels per fingertip ($6,100).
orcahandTactile sensing
Creates open-source anthropomorphic robotic hands for dexterous manipulation research. Self-dislocating joints prevent damage. Fully open-source under MIT/Creative Commons.
