每一个人形策略，每一个公开数据。
我们汇集了知名的人形策略与模型——共 25 个——以及它们所公布的真实、可溯源的数据。几乎每个数字都是各实验室在自己的任务套件上自报的，而且大多停留在仿真器中。目前还没有一套统一、独立的真实世界测量。Physical Turing 作为专注于人形机器人测试的公司，正在开展试验以补上这缺失的一列——下方每项指标都以白色显示实验室自报的数字，以琥珀色显示我们自己的真实世界测量，在试验落地之前显示为“Testing”。
已追踪 25 个人形策略
- 25
- 已追踪的人形策略
- 真实、可溯源的数据
- 9
- 仅在仿真中评估
- 没有真实硬件数据
- 16
- 公布了真实世界数据
- 大多为自定义任务
25 个策略自报Physical Turing
| 策略 | 本体 | 成功率 任务完成度 | 跟踪误差 动作保真度 | HumanoidBench 归一化得分 · 仿真 |
|---|---|---|---|---|
| ASAP Aligning Simulation and Real-World Physics for Agile Whole-Body Skills 全身控制真实世界 CMU · NVIDIA · 2025 | Unitree G1 | — Testing | 112 mmreal vs 159 mm baseline Testing | — Testing |
| HOVER Versatile Neural Whole-Body Controller for Humanoid Robots 全身控制真实世界 NVIDIA · CMU · 2024 | Unitree H1 | — Testing | 47.4 mmreal vs 51.0 mm specialist Testing | — Testing |
| HumanPlus Humanoid Shadowing and Imitation from Humans 全身控制真实世界 Stanford · 2024 | Unitree H1 (custom) | 60–100%real across 6 real skills Testing | — Testing | — Testing |
| HumanUP Learning Getting-Up Policies for Real-World Humanoid Robots 全身控制真实世界 UIUC · Simon Fraser · 2025 | Unitree G1 | 78.3%real getting up; vs 41.7% OEM Testing | — Testing | — Testing |
| HoST Learning Humanoid Standing-up Control across Diverse Postures 全身控制真实世界 Shanghai AI Lab · 2025 | Unitree G1 | 100%real 20/20, 4 terrains Testing | — Testing | — Testing |
| TWIST Teleoperated Whole-Body Imitation System 全身控制真实世界 Stanford · Simon Fraser · 2025 | Unitree G1 | — Testing | — Testing | — Testing |
| TrajBooster Boosting Humanoid Whole-Body Manipulation via Trajectory-Centric Learning 全身控制真实世界 OpenHelix · 2025 | Unitree G1 | up to 100%real best task; ~10 min data Testing | — Testing | — Testing |
| OmniH2O Universal and Dexterous Human-to-Humanoid Whole-Body Teleoperation and Learning 全身控制仿真 + 真实 CMU (LeCAR Lab) · 2024 | Unitree H1 | 94.1%sim AMASS imitation; + up to 10/10 real tasks Testing | — Testing | — Testing |
| ExBody2 Advanced Expressive Humanoid Whole-Body Control 全身控制真实世界 UC San Diego · 2024 | Unitree G1 | — Testing | 0.107 radreal mean per-joint; best vs baselines Testing | — Testing |
| H2O Learning Human-to-Humanoid Real-Time Whole-Body Teleoperation 全身控制仿真 CMU (LeCAR Lab) · 2024 | Unitree H1 | 72.5%sim AMASS; vs 85.5% oracle Testing | 167 mmsim global MPJPE Testing | — Testing |
| GR00T N1 An Open Foundation Model for Generalist Humanoid Robots 通用 VLA真实世界 NVIDIA · 2025 | Fourier GR-1 | 76.8%real +32% vs Diffusion Policy; 66.5% sim Testing | — Testing | — Testing |
| GR00T N1.5 Isaac GR00T N1.5 通用 VLA真实世界厂商自报 NVIDIA · 2025 | Unitree G1 / Fourier GR-1 | 98.8%real known objects; 84.2% novel Testing | — Testing | — Testing |
| Helix A Vision-Language-Action Model for Generalist Humanoid Control 通用 VLA真实世界厂商自报 Figure AI · 2025 | Figure 02 | 88.2→94.4%real barcode scan; vendor-reported Testing | — Testing | — Testing |
| EgoVLA Learning Vision-Language-Action Models from Egocentric Human Videos 通用 VLA仿真 UC San Diego · NVIDIA · 2025 | Unitree H1 (sim) | 77.8%sim 7 bimanual tasks Testing | — Testing | — Testing |
| iDP3 Generalizable Humanoid Manipulation with 3D Diffusion Policies 操作真实世界 Stanford · Simon Fraser · UPenn · 2024 | Fourier GR-1 | 9/10real unseen objects; vs 0–3/10 DP Testing | — Testing | — Testing |
| Berkeley Humanoid A Research Platform for Learning-based Control 运动真实世界 UC Berkeley · 2024 | Berkeley Humanoid | — Testing | 0.058 m/sreal vs 0.051 m/s sim Testing | — Testing |
| Humanoid-Gym Reinforcement Learning for Humanoid Robot with Zero-Shot Sim2Real Transfer 运动真实世界 RobotEra · Tsinghua · 2024 | RobotEra XBot-S / XBot-L | — Testing | — Testing | — Testing |
| TD-MPC2 Scalable, Robust World Models for Continuous Control RL 基线仿真 UC San Diego · 2023 | Unitree H1 (sim) | — Testing | — Testing | 0.710sim via SimbaV2 Testing |
| SimBa Simplicity Bias for Scaling Up Parameters in Deep RL RL 基线仿真 Sony AI · KAIST · 2024 | Unitree H1 (sim) | — Testing | — Testing | 0.606sim via SimbaV2 Testing |
| SimbaV2 Hyperspherical Normalization for Scalable Deep RL RL 基线仿真 KAIST · Sony AI · 2025 | Unitree H1 (sim) | — Testing | — Testing | 0.776sim low-UTD; best in class Testing |
| TDMPBC Self-Imitative Reinforcement Learning for Humanoid Robot Control RL 基线仿真 Westlake University · 2025 | Unitree H1 (sim) | 8/14 taskssim at 2M steps; vs 1 baseline Testing | — Testing | — Testing |
| FastTD3 Simple, Fast, and Capable Reinforcement Learning for Humanoid Control RL 基线仿真 + 真实 UC Berkeley · CMU · 2025 | Unitree H1 (sim) / Booster T1 | — Testing | — Testing | — Testing |
| DreamerV3 Mastering Diverse Domains through World Models (HumanoidBench baseline) RL 基线仿真 DeepMind algorithm · 2024 | Unitree H1 (sim) | — Testing | — Testing | 0.022sim via SimbaV2; high-UTD Testing |
| SAC Soft Actor-Critic (HumanoidBench baseline) RL 基线仿真 UC Berkeley algorithm · 2018 | Unitree H1 (sim) | — Testing | — Testing | 0.279sim via SimbaV2 Testing |
| PPO Proximal Policy Optimization (HumanoidBench baseline) RL 基线仿真 OpenAI algorithm · 2017 | Unitree H1 (sim) | — Testing | — Testing | — Testing |
由作者自报Physical Turing——独立的真实世界测量（Testing）real / sim 标示实验室在何处测量 · — = 未公布
HumanoidBench scores are normalized (0–1) on the low-UTD setting; the figures for TD-MPC2, SimBa, SAC and DreamerV3 are as compiled in the SimbaV2 benchmark (Lee et al., 2025), DreamerV3's at the high-UTD setting. Every other figure links to its own source via the policy name.
这些是他人的结果，由各实验室和厂商自行公布——在不同的机器人、不同的任务上，且大多在仿真中完成——因此彼此之间无法直接比较。Physical Turing 目前尚未在此发布任何属于自己的数字；每个琥珀色单元格都是我们正在开展的测量，而非主张。它们会随着试验的完成而逐步填入。
想要您的策略获得首个独立评分？
Physical Turing 开展真实世界试验，让您只交付真正就绪的成果。把您的策略或人形机器人告诉我们，我们将为您规划一次评估。
有问题想先了解？请发邮件至 support@physicalturing.ai。