KORTHOSCompaniesProductsVerified EventsSupply Chain

Product profile

Lumo-1

Lumo-1 is Astribot vision-language-action model for embodied robot control. It unifies reasoning and action so robots can interpret tasks, understand conte...

Facts

Company
Astribot
Product type
robotics_software
Status
available
Primary capability
VLA / embodied reasoning model
Website
https://arxiv.org/abs/2512.08580

About Lumo-1

Lumo-1 is Astribot's vision-language-action model for embodied robot control. It combines visual and language inputs with reasoning and action generation so a robot can interpret a task and produce manipulation behaviour.

The model sits between perception and the robot's actuators: it uses scene and instruction information to select actions, then supports closed-loop execution as the robot interacts with objects. Lumo-1 is a software research model for embodied control, with evaluation focused on how reasoning can be connected to physical robot action.

Market context

false, robot_control_policy, https://arxiv.org/abs/2512.08580, robot_foundation_model, no product image selected, Robotics Foundation Model, robot_reasoning_and_control, astribot_s1_and_cross_embodiment_data, vision_language_action_model, 2026-08-01, model_product, Robotics Foundation Model, Only a low-resolution official product mark was accessible; enhanced text plate rejected as not a real product image

Summary

Type: Robot AI

Core: VLA reasoning

Function: Control

Audience: Developers

Specifications

evaluation_mode
Real-world evaluations across challenging robotic tasks
robot_platforms_tested
Astribot S1
tasks_demonstrated
Long-horizon robot tasks responding to natural-language instructions
training_pipeline
Continued VLM pre-training, cross-embodiment robot co-training and action training with reasoning process

Operations

Product detail
false, robot_control_policy, https://arxiv.org/abs/2512.08580, robot_foundation_model, no product image selected, Robotics Foundation Model, robot_reasoning_and_control, astribot_s1_and_cross_embodiment_data, vision_language_action_model, 2026-08-01, model_product, Robotics Foundation Model, Only a low-resolution official product mark was accessible; enhanced text plate rejected as not a real product image

Research, Articles and recent verified events