LIBERO-Plus leaderboard

Robustness variant of LIBERO that perturbs camera, robot initial state, language, lighting, background, sensor noise and layout. Models trained on standard LIBERO are tested without further training.

Simulation · Franka Panda · MuJoCo (robosuite)Report an error
93.2%Best · InternW0-Δ
37Models ranked
OpenSaturation
Nov 18, 2026Numbers checked
#ModelTypeScoreDateSource
1InternW0-ΔShanghai AI LabWorld-action93.2%, self-reported by the authors, not a leaderboard entrySep 2026
2RoboHarnessHuawei Noah’s Ark LabAgent system92.4%, self-reported by the authors, not a leaderboard entryJul 2026
3Qwen-RobotManip-ContextAlibaba QwenVLA92.0%, self-reported by the authors, not a leaderboard entryJun 2026
4PDRVLAPeking University & JD.comVLA90.7%, self-reported by the authors, not a leaderboard entryOct 2026
5QuoVLATsinghua UniversityVLA90.3%, self-reported by the authors, not a leaderboard entryMay 2026
6Qwen-RobotManipAlibaba QwenVLA89.7%, self-reported by the authors, not a leaderboard entryJun 2026
7ACoT-VLAAGIBOTVLA87.5%, self-reported by the authors, not a leaderboard entryJan 2026
7RLDX-1RLWRLDVLA87.5%, self-reported by the authors, not a leaderboard entryMay 2026
9GAMKAIST AIPolicy86.3%, self-reported by the authors, not a leaderboard entryJun 2026
10InternVLA-A1.5Shanghai AI LabVLA85.8%, self-reported by the authors, not a leaderboard entryJul 2026
11RynnVLA-LatentAlibaba DAMO AcademyVLA85.6%, self-reported by the authors, not a leaderboard entryOct 2026
11π0.5Physical IntelligenceVLA85.6%, self-reported by the authors, not a leaderboard entryJun 2026
13ThinkWAMXi’an Jiaotong University et al.World-action84.8%, self-reported by the authors, not a leaderboard entryNov 2026
14ABot-M0.5AMAP CV LabWorld-action84.4%, self-reported by the authors, not a leaderboard entryJul 2026
15VLActStarVLA TeamVLA83.8%, self-reported by the authors, not a leaderboard entryAug 2026
16Cosmos PolicyNVIDIAWorld-action83.1%, self-reported by the authors, not a leaderboard entryJun 2026
16MachEmbodied-U0Li AutoWorld-action83.1%, self-reported by the authors, not a leaderboard entrySep 2026
18ImageWAMShanghai Jiao Tong UniversityWorld-action82.9%, self-reported by the authors, not a leaderboard entryJun 2026
19ResWAMMichigan State University et al.World-action82.6%, self-reported by the authors, not a leaderboard entryNov 2026
20Being-H0.7BeingBeyondVLA82.1%, self-reported by the authors, not a leaderboard entryMay 2026
21ABot-M0AMAP CV LabVLA81.6%, self-reported by the authors, not a leaderboard entryFeb 2026
22PatchWAMPatchWAM TeamWorld-action80.2%, self-reported by the authors, not a leaderboard entrySep 2026
23VLA-JEPAUSTCVLA79.5%, self-reported by the authors, not a leaderboard entryFeb 2026
24CSWAMMidea AIRCWorld-action78.5%, self-reported by the authors, not a leaderboard entrySep 2026
25StarVLAStarVLA TeamVLA76.9%, self-reported by the authors, not a leaderboard entryMar 3, 2026
26Elastic-WAMNational University of Singapore & RoboScienceWorld-action74.5%, self-reported by the authors, not a leaderboard entryNov 2026
27NF-VLAUniversity of Pennsylvania & ANUVLA74.2%, self-reported by the authors, not a leaderboard entryNov 2026
28GR00T N1.6NVIDIAVLA72.6%, self-reported by the authors, not a leaderboard entryMay 2026
29ReLAPBagel LabsVLA72.5%, self-reported by the authors, not a leaderboard entryNov 2026
30RelayWAMXinjiang University et al.World-action72.2%, self-reported by the authors, not a leaderboard entryNov 2026
31OpenVLA-OFTStanfordVLA71.4%, self-reported by the authors, not a leaderboard entryOct 2025
32LingBot-VARobbyantWorld-action69.5%, self-reported by the authors, not a leaderboard entryJul 2026
33GR00T N1.5NVIDIAVLA66.3%, self-reported by the authors, not a leaderboard entryMay 2026
34π0-FASTPhysical IntelligenceVLA62.5%, self-reported by the authors, not a leaderboard entryOct 2025
35π0Physical IntelligenceVLA56.1%, self-reported by the authors, not a leaderboard entryOct 2025
36NORASUTDVLA40.5%, self-reported by the authors, not a leaderboard entryOct 2025
37OpenVLAStanford / UC Berkeley et al.VLA16.3%, self-reported by the authors, not a leaderboard entryOct 2025

Metric: Zero-shot success on 10,030 perturbed LIBERO instances, mean of 7 perturbation dimensions. One result per model: RoboFYI re-run, then official leaderboard, then paper, then third party. Runs that deviate from the official protocol are listed on the model page but not ranked.

Wrong or missing number? Report an error