Robotics AIPilot deployment
Gemini Robotics
Google DeepMind · USA / UK · 2025
DeepMind's vision-language-action models that let robots think before they move.
Gemini Robotics brings Gemini's multimodal reasoning into the physical world: an embodied-reasoning model plans multi-step tasks while a VLA policy executes precise movements, including an on-device variant that runs entirely on the robot. Partnered with Apptronik and others, it can fold origami, pack lunchboxes and generalize to unseen objects.
#VLA model#embodied AI#on-device#Apptronik
Official website ↗▸Reviews
No reviews yet. Be the first!