Astribot introduces Lumo-1 model for embodied reasoning

Date: 2025-12-09 · Type: capability · Company: Astribot

Astribot introduced Lumo-1 as a generalist vision-language-action model designed to connect robot reasoning with robot action.

Astribot introduced Lumo-1 as a generalist vision-language-action model designed to connect robot reasoning with robot action.
The official research page and paper describe a three-stage training pipeline combining embodied reasoning, cross-embodiment robot data, and action training on Astribot S1 trajectories.
This is the most important software-side company event because it turns Astribot from a hardware demo story into a stronger embodied-intelligence platform story.