lab Google DeepMind Blog · 2h ago · src: 2d ago · releases

Google introduces Gemini Robotics 2 for whole-body control and multi-robot collaboration

from English

Google has introduced Gemini Robotics 2, an intelligence layer designed to provide adaptable robots with intelligent whole-body control, advanced dexterity, and the ability to collaborate in teams. This update expands physical AI capabilities beyond narrow, repetitive tasks by enabling robots to reason through movements and adapt to new bodies quickly.

Gemini Robotics 2: A vision-language-action model controlling full humanoids from feet to fingertips for complex manipulation.
Gemini Robotics ER 2: An embodied reasoning model that plans multi-step tasks lasting several minutes and enables multi-robot collaboration.
Gemini Robotics On-Device 2: An efficient model optimized for local execution, allowing adaptation to new robot embodiments in a few hours with minimal data.

The models are now available via Google AI Studio, Gemini Enterprise Agent Platform, and early-access partnerships. The update emphasizes safety through the ASIMOV-Agentic benchmark and enhanced proximity detection.

Importance 3/3 Trust 3/3 Google DeepMind Blog Google DeepMind AI agents Robotics

Read original

lab Google DeepMind Blog · 2h ago · 9 views

Google launches Gemini Robotics ER 2 with video understanding and multi-robot collaboration

Google has launched Gemini Robotics ER 2, a new embodied reasoning model designed to help robots think fast and plan multi-step tasks in real-time. The update introduces significant upgrades over the previous ER 1.6 version, including continuous video-based progress tracking, precision moment-finding, and native multi-robot collaboration capabilities.

arxiv arXiv cs.AI · 41d ago · 1 view

Finetuning VLA Models Requires Fewer Layers Than Thought

Vision-Language-Action models show severe layer-wise redundancy despite large parameter counts. A training-free compression method using Centered Kernel Alignment removes twin layers, reducing model depth by up to 50% and enabling 40-50% faster training and up to 30% faster inference without performance loss, validated across simulation and real-world robotic tasks.

arxiv arXiv cs.LG · 45d ago · 2 views

ROVE: Reinforcement Learning with Human Interventions for Humanoid Manipulation

ROVE enables humanoid Vision-Language-Action models to learn effective manipulation behaviors using imperfect human interventions. It combines a human-in-the-loop data collection pipeline with Optimistic Value Estimation and cross-embodiment supervision to prioritize high-value actions and improve robustness. ROVE outperforms baseline methods on real-world, contact-rich manipulation tasks through iterative rollout and intervention cycles.

lab Google — The Keyword (AI) · 2d ago · 5 views

Gemini API Managed Agents adds hooks, model selection, and free tier

Google has updated the Gemini Interactions API's Managed Agents to include environment hooks, explicit model selection, and free tier access. These capabilities build on previous releases introducing background tasks and remote MCP server integration.

lab Google DeepMind Blog · 3d ago · 10 views

Google releases Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber

Google has introduced three new models in its Flash series: Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber, designed to improve token efficiency, latency, and reliability for production AI agents.