Adroit Benchmark Leaderboard
In-hand manipulation benchmarks featuring the Adroit hand and diverse object tasks.
View Benchmark Repository25 models
Click row to expand details.
Sort by:
| Rank | Model | Mean Succ | Date | Details |
|---|---|---|---|---|
| 1 | GDP Two-Steps Diffusion Policy for Robotic Manipulation via Genetic Denoising | 97.20 | 2025.01 | |
| 2 | KOROL KOROL: Learning Visualizable Object Feature with Koopman Operator Rollout for Manipulation | 96.60 | 2024.07 | |
| 3 | DexHandDiff DexHandDiff: Interaction-aware Diffusion Planning for Adaptive Dexterous Manipulation | 92.50 | 2024.11 | |
| 4 | KODex On the Utility of Koopman Operator Theory in Learning Dexterous Manipulation Skills | 86.50 | 2023.03 | |
| 5 | KAN-We-Flow KAN We Flow? Advancing Robotic Manipulation with 3D Flow Matching via KAN & RWKV | 83.67 | 2026.02 | |
| 6 | BridgePolicy Sample from What You See: Visuomotor Policy Learning via Diffusion Bridge with Observation-Embedded Stochastic Differential Equation | 80.70 | 2025.12 | |
| 7 | ACT Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware | 78.95 | 2023.04 | |
| 8 | ManiFlow ManiFlow: A General Robot Manipulation Policy via Consistency Flow Training | 78.60 | 2025.09 | |
| 9 | K-UBM Going with the Flow: Koopman Behavioral Models as Pseudo Planners for Visuo-Motor Dexterity | 76.88 | 2026.02 | |
| 10 | MP1 MP1: MeanFlow Tames Policy Learning in 1-step for Robotic Manipulation | 75.67 | 2025.07 | |
| 11 | SAT Structural Action Transformer for 3D Dexterous Manipulation | 75.00 | 2026.03 | |
| 12 | ISS Policy ISS Policy : Scalable Diffusion Policy with Implicit Scene Supervision | 74.80 | 2025.12 | |
| 13 | FreqPolicy FreqPolicy: Frequency Autoregressive Visuomotor Policy with Continuous Tokens | 74.70 | 2025.06 | |
| 14 | SDM Policy Score and Distribution Matching Policy: Advanced Accelerated Visuomotor Policies via Matched Distillation | 74.00 | 2024.12 | |
| 15 | ManiCM ManiCM: Real-time 3D Diffusion Policy via Consistency Model for Robotic Manipulation | 72.30 | 2024.06 | |
| 16 | X-DP3 X-Imitator: Spatial-Aware Imitation Learning via Bidirectional Action-Pose Interaction | 71.17 | 2026.05 | |
| 17 | FlowPolicy FlowPolicy: Enabling Fast and Robust 3D Flow-based Policy via Consistency Flow Matching for Robot Manipulation | 70.30 | 2024.12 | |
| 18 | DP3 3D Diffusion Policy: Generalizable Visuomotor Policy Learning via Simple 3D Representations | 68.30 | 2024.03 | |
| 19 | Diffusion Policy Diffusion Policy: Visuomotor Policy Learning via Action Diffusion | 62.30 | 2023.03 | |
| 20 | RNR-DP Responsive Noise-Relaying Diffusion Policy: Responsive and Efficient Visuomotor Control | 57.40 | 2025.02 | |
| 21 | MambaPolicy MambaPolicy: Towards Efficient 3D Diffusion Policy with Hybrid Selective State Models | 52.30 | 2024.09 | |
| 22 | UniAct Universal Actions for Enhanced Embodied Foundation Models | 49.00 | 2025.01 | |
| 23 | HPT Scaling Proprioceptive-Visual Learning with Heterogeneous Pre-trained Transformers | 45.00 | 2024.09 | |
| 24 | Flow Matching Affordance-based Robot Manipulation with Flow Matching | 39.00 | 2024.09 | |
| 25 | AdaFlow AdaFlow: Imitation Learning with Variance-Adaptive Flow-Based Policies | 30.00 | 2024.02 |