Google DeepMind announced on July 30, 2026 that its updated Gemini Robotics 2 AI model can now control entire humanoid robots, expanding beyond the upper-body focus of its predecessor to support “whole-body motions” from feet to fingertips.
The new model enables humanoid robots to walk, crouch, stretch, and manipulate objects. Google shared videos demonstrating Apptronik’s Apollo 2 robot bending over to pick up a watering can and locating specific items on a shelf. Google DeepMind acknowledged that its robots “have more to advance in movement speed,” but described the update as “an important step towards the skills needed to complete more complex, real-world tasks that require whole-body coordination.”
Gemini Robotics 2 also adds improved dexterity through support for five-fingered hands, allowing robots to perform tasks such as sealing a Ziploc bag, tying a trash bag, and unscrewing a lightbulb.
Alongside the main model, Google DeepMind updated Gemini Robotics ER 2, a vision-language model that helps robots analyze surroundings, process instructions, and carry out multi-step tasks. The updated version handles longer task sequences and “now understands when tasks begin and end.” It also enables different robot types to collaborate — one video shows Apollo 2 directing Google’s dual-arm robot to sort tools into a bin while cleaning a garage.
Google DeepMind described Gemini Robotics ER 2 as its “safest robotics model to date,” noting it can better detect nearby humans, trigger safety responses, and bring a robot to a stop if someone approaches too closely.
The company also updated its Gemini Robotics On-Device Model, which runs locally without an internet connection. That model can now adapt more quickly to robots with different shapes, sensors, and degrees of freedom.
Source: The Verge