Google Deepmind's Gemini Robotics 2 takes over every robot body
google deepmind launched gemini robotics 2, a vla model controlling hardware from arms to humanoids, marking a push for universal robot intelligence.
google deepmind introduced gemini robotics 2, a vision-language-action model capable of controlling diverse robot hardware and managing full-body movement. the announcement also included gemini robotics er 2, an embodied reasoning model that replaces the april-released er 1.6 and is available via google ai studio. developers can request early access to the vla model now, signaling deepmind's intent to scale adaptive robotics beyond specific hardware constraints.