Dexhands

DEXTEROUS MANIPULATION

The world of dexterous hands.

Compare hardware. Discover applications. Share what works. An open reference and community for dexterous manipulation.

DexGraspBench

PKU / BODex team

A MuJoCo benchmark for comparing open-loop grasp poses and trajectories using explicit simulation metrics.

What you can explore

Compare grasp synthesis methods under a common task setup, including lifting and contact checks.

Before you use it

Simulation success and analytic metrics are distinct from real-hand reliability. The original baseline branch differs from current code. A reuse license was not identified in the checked root files.

Original contributors

Jiayi Chen, Yubin Ke and He Wang

License / access: Reuse terms unverified · benchmark assets retain their own licenses

Original code & documentation

FROM REFERENCE TO FIRST EXPERIMENT

Start with the right setup.

INPUTGrasps, object assets and evaluation settings
OUTPUTGrasp evaluation measurements
Environment
MuJoCo
Embodiment
Benchmark hand configuration

What you need

Follow the benchmark setup and obtain the separately referenced assets.

First useful check

Evaluate one documented example with its original settings.

Where it stops

Simulation results depend on assets, contact parameters and protocol; they are not physical validation. Reuse permissions require checking.

Documentation reviewed; not independently reproduced · reviewed 2026-09-24. Record the exact source revision when reproducing it.

Editorial reference · checked 2026-09-22

Start a discussion

Published posts are public. First contributions are reviewed. Only share material you have permission to disclose.

Report a post