Get Zlio on your home screen
One tap, full screen, like an app.
A major cross-robot VLA model from DeepMind pushes general-purpose robot control forward.
Gemini Robotics 2.5 is DeepMind's most capable vision-language-action model, turning sights and instructions into robot arm and hand movements across robot types. It reports 30-100% higher success and 20-30% faster tasks, with 96% in a Mercedes parts-kitting pilot.