Skip to main content
Rase AI Intelligence Newsletter

Robotics & physical AI Model or AI service

AgiBot’s GO-1 studies manipulation across more than one robot task

The AgiBot World research team describes Genie Operator-1 as a robot action model for manipulation. The authors use a learned representation of actions to train with demonstrations from several tasks. [1][2]

Model access
AgiBot World publishes a GO-1 model card and repository on Hugging Face. [2]

What the model does

The original AgiBot World Colosseo preprint introduced GO-1 as a policy trained with robot demonstrations. The team also published a GO-1 model repository. [1][2]

Why it matters

The project gives researchers a concrete way to study whether examples of different movements can help a robot perform a new manipulation task. [1]

Key dates

  1. AgiBot World Colosseo preprint describing GO-1 posted [1]

Ways to use it

The model card and research report are the starting points for matching the policy to the robot, camera and action interface required for a test. [1][2]

Model limitations

The results describe the team’s evaluation tasks and conditions. Other robots and use cases require validation. [1]

Sources and verification

Catalog evidence checked How we verify records

  1. Primary source Original AgiBot World Colosseo preprint describing GO-1
  2. Primary source agibot-world/GO-1

Related profiles

Robotics & physical AI

Gemini Robotics 2.5

Google DeepMind

The models are intended to interpret a robot's surroundings and support compatible robot tasks.

High confidenceEvidence checked

Robotics & physical AI

NEURA 4NE1

NEURA Robotics

A humanoid platform NEURA presents for development and industrial training, with multiple distinct hardware generations.

Medium confidenceEvidence checked

Robotics & physical AI

LUMO all-terrain humanoid

Shenzhen Dobot (DOBOT)

LUMO is a humanoid robotics platform whose proposed applications include interaction and service tasks.

Medium confidenceEvidence checked