Gemini Robotics 1.5: Advancing Robot Perception, Planning, and Tool Use Gemini Robotics 1.5 is a vision-language-action (VLA) model that enables robots to interpret visual input and instructions, translate them into motor commands, and perform multi-step tasks. The model emphasizes reasoning before action, providing transparency in task execution and supporting more reliable performance across different robotic platforms. Gemini Robotics-ER 1.5, a vision-language model (VLM), extends these capab