motus-robotics/Motus
Robotics Model
Motus is a vision-language-action model for bimanual manipulation, using flowmatching diffusion and latent-action modeling, built on the UniDiffuser architecture.
Catalog metadata
- Repository owner
- motus-robotics
- Source
- Hugging Face Models
- Task
- robotics
- Cameras
- 0
- License
- apache-2.0
Browse the catalog to compare robots, tasks, sensors, and licenses. Sign in to open protected download and model links.