Gemini Robotics ER 2 Now on Google Cloud: Enterprise Physical AI Takes a Major Step Forward

The Valley of the Sun, and all that is tasty therein
Post Reply
jiaolix
Posts: 1
Joined: Tue Oct 27, 2026 11:23 pm

Gemini Robotics ER 2 Now on Google Cloud: Enterprise Physical AI Takes a Major Step Forward

Post by jiaolix »

Google DeepMind's Gemini Robotics ER 2 — the embodied reasoning model that serves as a robot's high-level brain — is now available in public preview on the Gemini platform. The announcement brings frontier robotics intelligence into enterprise production environments for the first time, pairing it with Google Cloud's security, multi-agent orchestration, and end-to-end lifecycle management.

What Is Gemini Robotics ER 2?

Gemini Robotics ER 2 is a specialized Gemini model designed to help robots think, act, and adapt in the physical world. It combines spatio-temporal video understanding with long-horizon planning and built-in safety guardrails, working across any robot embodiment — from stationary arms to full humanoids. The model enables robots to converse naturally with humans, orchestrate low-level action models and hardware APIs, execute multi-minute tasks involving hundreds of decisions, and collaborate across multi-robot teams.

Key Capabilities for Enterprise Robotics

Physical Task Verification and Self-Correction
Robots need to confirm whether a step succeeded before proceeding. Gemini Robotics ER 2 reasons over continuous video feeds to classify task progress and pinpoint critical events — such as when to stop pouring a liquid — with 91.3% accuracy and a 0.96-second mean absolute frame distance. It runs at 4x the execution speed of many larger model categories at a fraction of the compute cost. The model also reads and interprets values across 10+ instrument categories, including circular dials, sight glasses, digital displays, and liquid thermometers.

One Intelligence Layer Across Any Embodiment

The model scales its intelligence from stationary arms and wheeled rovers to quadrupeds and complex humanoids. It also unlocks multi-robot collaboration, letting different robot types share semantic understanding of their workspace, divide labor, and hand off tasks.

Agentic Tool Orchestration and Natural Language Interaction

Developers can declare low-level control interfaces — VLA models, navigation APIs, or custom functions — as tools. Gemini Robotics ER 2 orchestrates these from natural language commands, allowing robots to explain their approach while acting, and letting operators redirect them without technical language.

Build, Simulate, and Deploy in Google Cloud

Simulate fast: Test robot behaviors in MuJoCo-Warp and NVIDIA Isaac Sim, generate synthetic scenarios, and auto-label workcell imagery.
Train smart: Curate and label demonstration videos in Cloud Storage and BigQuery, then train custom VLA models through the Gemini platform, GKE, or Cluster Director.
Operate easily: Deploy workflows with enterprise IAM and built-in function calling, offload latency-sensitive workloads to Google Distributed Cloud, and stream fleet telemetry into BigQuery.

How Customers Are Using It

Boston Dynamics sees video-based inspection as a new frontier for its Spot robot, moving beyond single-moment checks to continuous AI-driven analysis.
DeepReach AI uses ER 2 for accurate task understanding of recordings, accelerating data curation for model training.
Arm views advanced reasoning models as essential for helping robots interpret complex environments and plan actions.
ASTAR IMRE is giving lab robots the visual and physical flexibility to work with legacy infrastructure — reading touchscreens without APIs, loading centrifuges, and moving samples safely — edging closer to self-driving labs.

Accessing Gemini and Subscription Options

Developers and enterprise teams can start building with Gemini Robotics ER 2 today in preview via Model Garden. For individuals and teams looking to explore Gemini's broader capabilities, z2u.com offers a convenient marketplace with flexible options, including Gemini AI Subscription plans, the ability to Buy Gemini Advanced for access to premium features, and Buy Gemini Subscription Key listings for quick activation. Whether you're prototyping robotics workflows or upgrading your AI toolkit, z2u.com provides accessible entry points to the Gemini ecosystem.

Get Started

Build in Gemini platform: Visit the Gemini Robotics ER 2 documentation and pricing page.
Explore developer guides: Check out sample code and bidirectional streaming guides in the Gemini API Robotics documentation.
Read the research: Dive into benchmarks and the safety technical report in the Google DeepMind announcement.
Explore Physical AI on Google Cloud: Visit cloud.google.com/solutions/physical-ai to learn how Google Cloud infrastructure supports simulation, training, and deployment.
gemini2.jpg
gemini2.jpg (21.79 KiB) Viewed 83 times
Post Reply