Comparable Metrics
Benchmarks are grouped for apples-to-apples performance checks.
Robot Learning Benchmarks
Standardized evaluation for robot manipulation — RLBench, LIBERO, CALVIN, and more. Success rates, task completion, evaluation metrics.
Robotics Benchmark LeaderboardsLive
Real-time rankings from Papers with Code, updated daily across manipulation, locomotion, and navigation tasks.
Loading leaderboard data…
Quick Browse
Popular Categories
Fast Tags
Popular Tags
Benchmarks for Manipulation
5 benchmarks match current filters.
Simulation
RLBench
Simulationmanipulationtask-suite
100+ manipulation tasks in PyRep. Widely used for VLA evaluation. BridgeVLA 88.2%, InternVLA 95%+ on subsets.
View benchmark →SimulationLIBERO
Simulationlifelong-learningmanipulation
Lifelong learning benchmark. 130 tasks, spatial/object/goal suites. RoboSuite. 95.9% SOTA (InternVLA).
View benchmark →SimulationCALVIN
Simulationlanguage-conditionedlong-horizon
Composing Actions from Language and Vision. Long-horizon, language-conditioned. RoboFlamingo strong baseline.
View benchmark →Real RobotGoogle Robot Benchmark
Real Robotreal-worldmulti-task
Real-world manipulation. 700+ tasks. WidowX, various embodiments. Success rate, multi-task evaluation.
View benchmark →Real RobotCOLOSSEUM
Real Robotreal-worldscale
Large-scale real-robot benchmark. Diverse tasks, environments. BridgeVLA 64%.
View benchmark →Suggested Models & Datasets
Suggested stack based on selected benchmark type.
Real vs Sim Coverage
Evaluate both controlled and deployment-oriented settings.
Model Mapping
Each benchmark path links to compatible model families.
Execution Support
Support for data capture and evaluation operations when needed.
Need Evaluation or Data for Your Benchmark?
We provide data collection and real-world evaluation support.
Continue the loop
Benchmarks compare policies in sim; our eval loop scores them on physical rigs.