NewsBite analysis · AI, Robotics
MarkTechPost
Google DeepMind Ships Three Physical AI Models For Whole Body Control, Dexterity And Multi Robot Collaboration
Google DeepMind has announced the release of Gemini Robotics 2, a suite of three AI models that significantly enhance robotic capabilities. Designed for whole body control, these models allow for multi-robot collaboration and dexterous manipulation. This advancement marks a shift from pre-programmed systems, enabling robots to adapt to dynamic environments. Among the models, Gemini Robotics ER 2 showcases embodied reasoning, processing multimodal inputs and executing complex tasks with high accuracy.
Act nowConfidence: HighImmediate horizon
Published 30 Jul 2026, 18:40 UTC5 min read
Visit original article ↗Lead image — MarkTechPost
Summary
TL;DR
- Launch of Gemini Robotics 2 with three advanced AI models
- Improvements in whole body control and dexterity for robots
- New AI models include a vision-language-action model
- Multi-robot teamwork capabilities introduced
- ER 2 model achieves high accuracy in task progression detection
Interesting facts
- Gemini Robotics 2 can control humanoids and multi-arm robots.
- The safety benchmark ASIMOV-Agentic is now publicly available.
- Robots can adapt to different physical embodiments quickly.
- The model achieves notable accuracy against previous generations.