Google DeepMind has introduced Gemini Robotics-ER 1.6, an update to its embodied reasoning model line. The new release is designed to help autonomous robots better interpret and act in the physical world.
The model emphasizes two related capabilities: spatial reasoning and multi-view understanding. These are central to real-world robotics, where robots need to perceive scenes from different angles and reason about objects and spaces reliably.
Since this announcement comes from a single source, there are no competing reports to compare. The details reflect DeepMind’s own description of the system’s intended capabilities.