EN

MODEL CATALOG

Embodied AI models

Compare released model versions, checkpoints, embodiment compatibility, and benchmark evidence.

All Vision-language-action Robot policy World model

NVIDIA Isaac GR00T

4 Versions

GR00T N1.5N1.5

GR00T N1.6N1.6

GR00T N1.7N1.7

Figure Helix

1 Versions

1X World Model

1 Versions

UBTECH Co-Agent

1 Versions

BrainNet 2.0 + Co-AgentBrainNet 2.0 + Co-Agent

TOPReward

1 Versions

LeRobot TOPRewardLeRobot TOPReward

Robometer

1 Versions

Robometer 4BRobometer 4B

Default checkpoint: robometer/Robometer-4B

SARM

1 Versions

LeRobot SARMLeRobot SARM

WALL-OSS

Vision-language-action

1 Versions

WALL-OSS FlowWALL-OSS Flow

Default checkpoint: x-square-robot/wall-oss-flow

Multitask DiT

Robot policy

1 Versions

LeRobot Multitask DiTLeRobot Multitask DiT

EVO1

Vision-language-action

1 Versions

FastWAM

Robot policyWorld model

1 Versions

FastWAM baseFastWAM base

Default checkpoint: lerobot/fastwam_base

LingBot-VA

Vision-language-actionWorld model

1 Versions

LingBot-VA baseLingBot-VA base

Default checkpoint: lerobot/lingbot_va_base

EO-1

Vision-language-action

1 Versions

VLA-JEPA

Vision-language-actionWorld model

1 Versions

VLA-JEPAVLA-JEPA

Default checkpoint: lerobot/VLA-JEPA-Pretrain

ACT

Robot policy

1 Versions

LeRobot ACTLeRobot ACT

MolmoAct2

Vision-language-action

2 Versions

MolmoAct2-ThinkMolmoAct2-Think

Default checkpoint: allenai/MolmoAct2-Think

MolmoAct2MolmoAct2

Default checkpoint: allenai/MolmoAct2

RDT

Vision-language-action

1 Versions

RDT-1BRDT-1B

Default checkpoint: robotics-diffusion-transformer/rdt-1b

Octo

Vision-language-action

2 Versions

Octo-Base 1.5Octo-Base 1.5

Default checkpoint: rail-berkeley/octo-base-1.5

Octo-Small 1.5Octo-Small 1.5

Default checkpoint: rail-berkeley/octo-small-1.5

X-VLA

Vision-language-action

1 Versions

X-VLA 0.9BX-VLA 0.9B

Default checkpoint: lerobot/xvla-base

SmolVLA

Vision-language-action

1 Versions

SmolVLA baseSmolVLA base

Default checkpoint: lerobot/smolvla_base

π₀

Vision-language-action

3 Versions

π₀.₅π₀.₅

Default checkpoint: openpi π₀.₅ base

π₀-FASTπ₀-FAST

Default checkpoint: lerobot/pi0fast-base

π₀ baseπ₀ base

Default checkpoint: lerobot/pi0_base

GR00T

Vision-language-actionRobot policy

2 Versions

GR00T N1.7GR00T N1.7

Default checkpoint: nvidia/GR00T-N1.7-3B

GR00T N1.5GR00T N1.5

Default checkpoint: nvidia/GR00T-N1.5-3B

OpenVLA

Vision-language-action

1 Versions

OpenVLA 7BOpenVLA 7B

Default checkpoint: openvla/openvla-7b

UnifoLM-VLA

1 Versions

UnifoLM-VLA-0UnifoLM-VLA-0

Submit a source or correction