Skip to main content
Rase AI Intelligence Newsletter

Robotics & physical AI Model or AI service

Unitree’s UnifoLM-WMA-0 links scene prediction with robot actions

Unitree’s UnifoLM-WMA-0 is a research system for predicting interactions between a robot and its surroundings and selecting movements for manipulation. [1]

Model access
Unitree publishes code on GitHub and links to separately hosted model checkpoints. [1]

What the model does

Unitree documents a world model that predicts interactions, and a decision mode for choosing actions. Its repository announced release of code and model weights in September 2025. [1]

Why it matters

Learning to predict what may happen after an action gives researchers another way to study robot control alongside direct action prediction. [1]

Key dates

  1. Unitree announced code and model weight release [1]

Ways to use it

Unitree documents separate checkpoints and configurations for research. Match the checkpoint to the robot and task before evaluating performance. [1]

Model limitations

Unitree’s examples use selected research platforms and tasks. Performance on other robots and in new environments requires validation. [1]

Sources and verification

Catalog evidence checked How we verify records

  1. Primary source Unitree’s UnifoLM world-model-action repository

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