EmbodiedCLUE-VLA RoboTwin Evaluation
Robot policies perform ten two-handed manipulation tasks under randomized RoboTwin 2.0 Hard conditions. The score averages task success rates over twenty episodes per task.
SuperCLUE publishes scores for 17 models in its 2026 results workbook.
- Made by
- SuperCLUE
- Released
- August 2026
- Kind
- Simulated robots
- Level
- Everyday
- Scored
- Checked automatically
- Answers
- Agent task
- Measures
- Capability
- Interaction
- Agentic
- Environment
- RoboTwin
- Items from
- Generated
- Refresh
- One fixed set
- Run by
- A host
- Based on
- A version of another benchmark
Hy-Embodied-0.5-VLA-RoboTwin tops the leaderboard with 73%
Best score per entry from the benchmark's own leaderboard, read on November 21st, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders.
- 73.0 Hy-Embodied-0.5-VLA-RoboTwin
- 68.5 Motus
- 61.0 InternVLA-A1.5-RoboTwin
- 57.0 ME-Dex-1.0-RoboTwin (VAE)
- 48.5 OpenWAM-Alpha-Sim-RoboTwin
- 47.5 UniWAM-robotwin
- 45.0 Fast-VLA
- 40.0 WorldScape Policy 2.0
- 39.0 LingBot-VLA-V2-6B-RoboTwin
- 35.0 PICA-WAM
- 34.5 ABot-M0-RoboTwin2
- 34.0 StarVLA Qwen3-VL-OFT
- 30.5 GigaBrain-0.7-3.5B-RoboTwin2.0
- 30.5 InternW0-Delta-RoboTwin
- 20.5 PrismWAM
- 13.0 π0.5-RoboTwin2
- 10.0 X-VLA-0.9B
- 1 Hy-Embodied-0.5-VLA-RoboTwin Full 73%
- 2 Motus Full 68.5%
- 3 InternVLA-A1.5-RoboTwin Full 61%
- 4 ME-Dex-1.0-RoboTwin (VAE) Clean 57%
- 5 OpenWAM-Alpha-Sim-RoboTwin Full 48.5%
- 6 UniWAM-robotwin Clean 47.5%
- 7 Fast-VLA Full 45%
- 8 WorldScape Policy 2.0 Full 40%
- 9 LingBot-VLA-V2-6B-RoboTwin Full 39%
- 10 PICA-WAM Clean 35%
- 11 ABot-M0-RoboTwin2 Full 34.5%
- 12 StarVLA Qwen3-VL-OFT Full 34%
- 13 GigaBrain-0.7-3.5B-RoboTwin2.0 Clean 30.5%
- 14 InternW0-Delta-RoboTwin Clean 30.5%
- 15 PrismWAM Clean 20.5%
- 16 π0.5-RoboTwin2 Full 13%
- 17 X-VLA-0.9B Clean 10%
Source: Official EmbodiedCLUE-VLA RoboTwin Evaluation leaderboard
| # | Entry | Score ± | Setting | Submitted by | Date | Source |
|---|---|---|---|---|---|---|
| 1 | Hy-Embodied-0.5-VLA-RoboTwin | 73% ±5 | Full | 腾讯混元 | Official leaderboard | |
| 2 | Motus | 68.5% ±5.5 | Full | 清华大学/生数科技 | Official leaderboard | |
| 3 | InternVLA-A1.5-RoboTwin | 61% ±5 | Full | 上海AI Lab | Official leaderboard | |
| 4 | ME-Dex-1.0-RoboTwin (VAE) | 57% ±5.5 | Clean | 理想汽车 | Official leaderboard | |
| 5 | OpenWAM-Alpha-Sim-RoboTwin | 48.5% ±4.5 | Full | 新加坡国立大学/清华大学/北京大学等 | Official leaderboard | |
| 6 | UniWAM-robotwin | 47.5% ±5.5 | Clean | OLA/香港科技大学(广州)等 | Official leaderboard | |
| 7 | Fast-VLA | 45% ±4 | Full | 华为技术有限公司/华为中央研究院 | Official leaderboard | |
| 8 | WorldScape Policy 2.0 | 40% ±4.5 | Full | Manifold AI(流形空间) | Official leaderboard | |
| 9 | LingBot-VLA-V2-6B-RoboTwin | 39% ±5.5 | Full | 蚂蚁灵波/Robbyant | Official leaderboard | |
| 10 | PICA-WAM | 35% ±4.5 | Clean | 上海交通大学/上海AI Lab/复旦大学 | Official leaderboard | |
| 11 | ABot-M0-RoboTwin2 | 34.5% ±5.5 | Full | 阿里高德/AMAP-CV Lab | Official leaderboard | |
| 12 | StarVLA Qwen3-VL-OFT | 34% ±4.5 | Full | StarVLA团队 | Official leaderboard | |
| 13 | GigaBrain-0.7-3.5B-RoboTwin2.0 | 30.5% ±5 | Clean | GigaAI(极佳科技) | Official leaderboard | |
| 14 | InternW0-Delta-RoboTwin | 30.5% 26 to 35.5 | Clean | 上海AI Lab | Official leaderboard | |
| 15 | PrismWAM | 20.5% ±4 | Clean | 清华大学/NVIDIA | Official leaderboard | |
| 16 | π0.5-RoboTwin2 | 13% 9.5 to 17 | Full | Physical Intelligence | Official leaderboard | |
| 17 | X-VLA-0.9B | 10% ±3.5 | Clean | 清华AIR等 | Official leaderboard |
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Robotics
- ERQA In use 400 multiple-choice visual questions from robot and egocentric footage on spatial, trajectory, action and state reasoning. Accuracy.
Gemini Robotics ER 2 78.5%
- SimplerEnv In use Real-to-sim replicas of Google Robot and WidowX/Bridge setups. Policies scored by success rate on pick, move, drawer and Bridge tasks. Praxis-1 80%
- RoboLab In use Zero-shot DROID Franka policies in Isaac Lab on 120 tabletop tasks at three instruction levels, 10 episodes each. Success rate. Praxis-1 46.2%
- RoboDojo In use Generalist manipulation policies run 42 Isaac Sim tasks and 18 remote real-robot tasks. Scored by task score and success rate. GPT-6 Astra 27.6%
- Progress classification (Google DeepMind) In use Models assign each frame of a robot video to one of five task-progress bands from 0 to 100 percent. Scored by accuracy.
Gemini Robotics ER 2 57.4%
- Human proximity (Google DeepMind robotics) In use Models judge from robot camera views whether a human is within 1 meter, a spatial awareness check for safe robot behavior. Scored by accuracy.
Gemini Robotics ER 2 93%
Data audit
7 of 8 checks pass, 1 warning. Look first at the Task access check. We don’t know whether the tasks are public. Every benchmark
Catalog record 1 warning 1 pass
- Warning Task access We don’t know whether the tasks are public.
- Pass Confidence Checked against a primary source.
Scores 2 pass 3 notes
- Pass Independent runs Scores from the official leaderboard (17 entries).
- Pass Models scored 17 entries on its leaderboard, none a model we track.
- Note Newest score No scores yet.
- Note Frontier models No models scored yet.
- Note Headroom No scores yet.
Leaderboard import 3 pass
- Pass Importer We read the board on superclueai.com: 17 entries at the last read, none a model we track.
- Pass Last read November 21st, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.