Google DeepMind's Gemini Robotics 2.5 hits 96% success sorting metal parts

google deepmind unveiled gemini robotics 2.5, a vision-language-action model that processes visual information and translates it into robot movements. in a pilot test on mercedes-benz production lines in hungary and germany, apptronik's humanoid robot apollo 2 sorted metal parts with a 96 percent success rate. the model achieves a 30 to 100 percent higher success rate and completes tasks 20 to 30 percent faster than previous baselines. google deepmind also released the open-source benchmark asimov-jailbreak to test the safety of robotics models.