Google DeepMind has introduced Gemini Robotics 2, an AI system designed for advanced robot control. This new system enables robots to perform whole-body movements, execute dexterous five-finger manipulations, and collaborate in multi-robot teams, moving beyond previous limitations of pre-programmed or tele-operated robots. Gemini Robotics 2 is comprised of three distinct models: a vision-language-action (VLA) model for motor control, an embodied reasoning (ER) model for high-level planning and understanding, and an on-device VLA model optimized for local execution. These models have been demonstrated on robots like Apptronik's Apollo 2, showcasing capabilities such as complex object manipulation and navigation. AI
IMPACT Enables more adaptable and collaborative robots, pushing the boundaries of embodied AI and dexterous manipulation.
RANK_REASON Frontier-lab model release with system card [lever_c_demoted from frontier_release: ic=1 ai=1.0]
- Apptronik Apollo 2
- ASIMOV-Agentic
- Gemini 3.5 Flash
- Gemini Robotics 1.5
- Gemini Robotics 2
- Gemini Robotics ER 2
- Gemini Robotics On-Device 2
- Gemma
- Google DeepMind
- Hugging Face
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
