Repository
THE EMBODIED AI DIRECTORY
Benchmarks
Explore tasks and evaluation suites for manipulation, navigation and control.
Repository
Repository
Repository
Repository
Repository
actionshift · Archerkattri
ActionShift evaluates adaptation to hidden robot action-interface contracts on ManiSkill tasks.
Manipulation
Repository
RMBench · RoboTwin-Platform
Memory-Dependent Manipulation Benchmark based on RoboTwin
Manipulation
Repository
Repository
dexart-release · Kami-code
DexArt: Benchmarking Generalizable Dexterous Manipulation with Articulated Objects, CVPR 2023
Manipulation
Repository
Meta-World
A manipulation benchmark for multi-task and meta-learning experiments.
Manipulation
Repository
vla-evaluation-harness · allenai
One framework to evaluate any VLA model on any robot simulation benchmark.
Policy evaluation
Repository
softgym · Xingyu-Lin
SoftGym provides benchmark environments for deformable-object manipulation.
Manipulation
Repository
lidar_slam_ros2 · rsasaki0109
ROS 2 LiDAR SLAM for pointcloud-map authoring, benchmarking, and Autoware-compatible map workflows.
Navigation & embodied interaction
Repository
actprobe · air-embodied-brain
ActProbe provides probes and benchmarks for early failure detection in generative robot policies.
Policy evaluation
Website
HumanoidBench
Whole-body control tasks for evaluating humanoid robot learning.
Locomotion & whole-body control
Repository
robot colosseum · robot-colosseum
Robot Colosseum benchmarks generalization in robotic manipulation.
Manipulation
Repository
LocoMuJoCo
A benchmark for learning and evaluating locomotion behaviors.
Locomotion & whole-body control
Repository
furniture bench · clvrai
FurnitureBench provides real-world furniture-assembly benchmarks and related simulation resources.
Manipulation
Repository
Gymnasium Robotics
Robotics reinforcement-learning environments using the Gymnasium interface.
Manipulation
Repository
Repository
world-model-eval · world-model-eval
Code for "Evaluating Robot Policies in a World Model".
Policy evaluation
Website
ALFRED
A benchmark for following language instructions in interactive environments.
Navigation & embodied interaction
Repository
bark · bark-simulator
BARK provides simulation and benchmarking tools for autonomous-driving behavior planning.
Navigation & embodied interaction
Website
BEHAVIOR
Household activity benchmarks for embodied AI and robotic behavior.
Navigation & embodied interaction
Repository
Bench2Dex · Bench2Dex
A Bimanual Manipulation Benchmark across Dexterous Hands with Simulated Visuo-Tactile Support
ManipulationThe ecosystem keeps growing.
Know a resource that belongs here?