Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

EBench

Benchmarks

Task NameDataset NameSOTA ResultTrend
Visual Quality EvaluationEBench-18K
SRCC0.9466
44
Instruction Alignment EvaluationEBench 18K
SRCC92.09
22
Mobile manipulationEBench (test)
Success Rate (SR)35.2
6
Mobile manipulationEBench Unseen (val)
Success Rate32.8
6
Mobile manipulationEBench (val-train)
Success Rate (SR)43.1
6
Robot ManipulationEBench Laboratory 2026 (Overall)
SR45.6
6
Robot ManipulationEBench Long Horizon Laboratory 2026
Success Rate (SR)29.9
6
Robot ManipulationEBench Simple PnP Laboratory 2026
Success Rate (SR)56.5
6
Robot ManipulationEBench Table Top Laboratory 2026
Success Rate50
6
Showing 9 of 9 rows