Gemini Robotics 2: Google DeepMind's Humanoid Robot AI
Blog News5 min read # Google DeepMind's Gemini Robotics 2 Can Control a Humanoid From Feet to Fingertips DeepMind's new vision-language-action model drives a full humanoid, legs included, with Apptronik's Apollo 2 as the demonstration platform.
Key takeaways
- The most recent developments show that humanoid robots are becoming both a commercial focus and a geopolitical flashpoint.
- In the United States, the federal government announced a ban on imports of foreign‑made humanoid and quadruped robots, specifically targeting Chinese manufacturers on national‑security grounds, a move that will cut off the dominant Chinese market share—estimated at roughly 85%—from its largest export market (U.S. ban reported in late July 2026).
- At the same time, major U.S. airlines have tightened safety rules: Delta Air Lines and United Airlines have joined Southwest in prohibiting human‑ or animal‑like robotic devices from cabin or checked‑baggage areas, citing battery‑safety concerns (policy updates issued in June and July 2026).
- In China, electric‑vehicle giant BYD confirmed it will unveil its first commercial humanoid robot in August at its Di Space experience centres, planning to place two or three robots in every BYD store within the next few years (announcement made July 28‑29, 2026).
- Meanwhile, Google DeepMind released Gemini Robotics 2, a vision‑language‑action model that can control an entire humanoid body—including walking, balancing and dexterous hand manipulation—using Apptronik’s Apollo 2 platform, highlighting a shift toward foundation‑model intelligence across diverse robot hardware (demo unveiled July 30, 2026).
Blog
News5 min read
Google DeepMind's Gemini Robotics 2 Can Control a Humanoid From Feet to Fingertips
DeepMind's new vision-language-action model drives a full humanoid, legs included, with Apptronik's Apollo 2 as the demonstration platform. Here is what it does, what the published success rates say, and why the strategy matters more than the demo.
Radica Maneva, Robozaps
Radica Maneva
Last fact-checked
On this page
Google DeepMind released Gemini Robotics 2 on 30 July 2026, a vision-language-action model that controls an entire humanoid robot: walking, balancing, crouching, and handling objects with five-fingered hands. Apptronik's Apollo 2 is the demonstration platform. ## Sources & references
- Gemini Robotics 2 brings whole body intelligence to robots Google DeepMind · accessed Aug 1, 2026
- Boston Dynamics & Google DeepMind Form New AI Partnership to Bring Foundational Intelligence to Humanoid Robots Boston Dynamics · accessed Aug 1, 2026
- Welcome to Robot Park: Where Apptronik's Apollo Goes to Work Training the Next Generation of Humanoid Robot Intelligence Apptronik · accessed Aug 1, 2026 ## Why this matters more than the demo
Humanoid hardware is fragmenting. Apptronik, Figure, Tesla, Unitree, Boston Dynamics, AgiBot and a dozen others are all building bodies. Foundation models are consolidating instead, because training them takes data and compute that most robot makers cannot fund alone.
The result is an Android-shaped possibility: many robot bodies, a small number of intelligence suppliers underneath them. Google already demonstrates Gemini across humanoids, two-armed systems, multi-finger hands and grippers, and it signed an AI partnership with Boston Dynamics in January 2026 covering the next-generation Atlas.
New robots, funding, and signals. Every Thursday.
Unsubscribe anytime.
Mentioned in this article