Physical AI
NVIDIA Omniverse
Build simulation-ready worlds for physical AI by integrating modular, agent-ready libraries into existing 3D applications and workflows.
Overview
What is NVIDIA Omniverse?
Physical AI systems need simulation-ready worlds for design, training, testing, and validation before real-world deployment. But a 3D scene can look complete and still lack the structure and physical properties required for simulation.
NVIDIA Omniverse™ libraries, part of NVIDIA Agent Toolkit, are modular software components that developers and AI agents can integrate into existing applications and workflows to prepare and validate simulation-ready worlds. Frontier AI agents can use these to prepare 3D content, assemble and test simulation workflows, and flag issues for developer review. This allows software makers to make their trusted solvers, simulators, tools, and data available to agents while retaining control of their applications, workflows, and customer experience.
Technology
Core Omniverse Libraries and Services
Connect and Structure 3D Data
Use OpenUSD to map native application data into structured 3D scenes that can move across applications and simulation pipelines. Explore Python and C++ samples for authoring and exchanging USD data.
Check Asset Readiness
Use SimReady to check OpenUSD assets against simulation-ready requirements before downstream training, testing, or deployment. Review which requirements pass or fail and identify asset properties that need attention.
Render 3D Worlds for Simulation
Use ovrtx to bring NVIDIA RTX rendering and sensor simulation into existing applications. Create interactive visualizations and generate camera, lidar, radar, and other sensor outputs from OpenUSD scenes.
Simulate Physical Behavior
Use ovphysx to bring GPU-accelerated physics into existing applications. Simulate motion, contacts, and collisions for robotics and digital twin workflows, with Python and C/C++ examples to get started.
Starting Options
Get Started With Omniverse
Use Cases
How Teams Use Omniverse for Physical AI
-
AI Factory
Digital Twins -
Industrial
Digital Twins -
Robot Simulation
and Learning -
Simulation-Ready
3D Data -
Synthetic Data
Generation -
Autonomous Vehicle
Simulation
AI Factory Digital Twins
Optimize AI factory performance—from design and construction through operations.
Connect compute, networking, power, and cooling in a digital twin to test AI factory performance before construction and optimize throughout operations.
- Connect facility and equipment data with OpenUSD and SimReady assets.
- Optimize layouts, power, and cooling with the NVIDIA Omniverse DSX Blueprint.
- Test design and operational changes in simulation before implementation.
Industrial Digital Twins
Build simulation-ready industrial facility digital twins to test and optimize operations.
Prepare industrial environments for testing, optimizing, and validating physical AI workflows before real-world deployment.
- Connect OpenUSD scene data across design, engineering, and simulation workflows.
- Add RTX rendering, physics, sensors, and validation to make facilities simulation-ready.
- Inspect environments, surface issues, and return evidence for review with agent-ready tools.
KION
Robot Simulation and Learning
Train robot policies and test behavior in simulation before real-world deployment.
Bring physics, sensor simulation, and OpenUSD into robotics testing and learning workflows with Omniverse libraries.
- Prepare OpenUSD scenes with materials, physics properties, sensors, and semantic data.
- Test robot perception, motion, and interactions using physics and sensor simulation.
- Create varied tasks and environments to train and evaluate robot policies.
Simulation-Ready 3D Data
Prepare and validate 3D assets to build simulation worlds for physical AI.
Build physically accurate assets that 3D developers and AI agents can use to create simulation-ready worlds.
- Convert existing 3D content into OpenUSD assets for simulation workflows.
- Add physics properties, materials, and semantic labels required by the intended use case.
- Validate assets against SimReady requirements and reuse content across supported tools and runtimes.
Synthetic Data Generation
Generate physics-based sensor data to train and test physical AI models.
Generate physically grounded synthetic data from simulation-ready worlds to train and test physical AI models.
- Configure scene variation, sensor layouts, lighting, materials, and rendering outputs.
- Produce synthetic data with physically based rendering and sensor simulation.
- Trigger data-generation tasks, inspect outputs, and flag issues for review.
Autonomous Vehicle Simulation
Build virtual driving environments to test AV perception and sensor responses.
Prepare simulation-ready environments for testing perception systems, sensors, and autonomous vehicle scenarios.
- Structure road, facility, and scenario data with sensor-ready OpenUSD assets.
- Simulate cameras, lidar, radar, and visual preflight for perception workflows.
- Inspect sensor setup and flag readiness issues before downstream testing.
Success Stories
Physical AI Workflows in Practice
Ecosystem
Omniverse Across the Physical AI Ecosystem
See how partners and developers are connecting Omniverse capabilities into applications and workflows for physical AI, digital twins, robotics, and simulation.
Resources
The Latest From Omniverse
Frequently Asked Questions
NVIDIA Omniverse provides libraries, APIs, services, and agent-ready tools for building physical AI applications and simulation-ready 3D workflows. Developers use Omniverse to integrate OpenUSD data interoperability, RTX rendering, physics, sensor simulation, validation, and runtime capabilities into existing applications and workflows.
Developers can build physical AI simulation applications, industrial digital twins, robotics simulation workflows, synthetic data pipelines, and tools for preparing 3D worlds for simulation. Omniverse helps teams connect 3D data, simulate sensors and physics, validate assets, and move from visual 3D content toward simulation-ready environments.
Frontier AI agents—such as those powered by OpenAI GPT‑6 Astra—can become an entry point to NVIDIA Omniverse libraries, using them as callable tools and skills to create and prepare 3D content, assemble simulation workflows, run experiments, and validate results. ISVs can integrate these capabilities into existing applications, build agents into their products, or make trusted solvers, simulators, tools, and data available for agents to call—while retaining control of the customer experience, workflows, and key decisions.
Yes. Omniverse libraries and services help developers and software makers add simulation capabilities into existing tools, services, and workflows. This lets teams use Omniverse for rendering, physics, sensor simulation, OpenUSD interoperability, and validation without replacing their differentiated applications or user experiences.
Start with Omniverse libraries when you want to integrate rendering, physics, sensor simulation, OpenUSD operations, or validation into an application or service. Use agent skills when you want AI agents to call Omniverse tools for defined tasks. Start from blueprints when you want reference workflows that show how NVIDIA technologies fit together for physical AI development.
As of May 2026, Omniverse is freely available for both development and production use with no NVIDIA AI Enterprise subscription required. Software built with Omniverse may also be freely redistributed under the same terms. This aligns with NVIDIA’s established approach to offering other libraries for no cost.
Use of all Omniverse software is subject to the NVIDIA End User License Agreement (EULA) and Product-Specific Terms.
When Omniverse is integrated without an NVIDIA AI Enterprise subscription, support is limited to community support via the NVIDIA Developer Forum and Discord.
An NVIDIA AI Enterprise subscription can be purchased from an NPN reseller or a Cloud Service Provider (CSP) Marketplace for customers that require Enterprise Support.
NVIDIA Cosmos provides open world foundation models and frameworks that help physical AI systems reason, generate world states, and learn actions. Omniverse provides the simulation technologies developers use to build, render, and test structured 3D environments.
Developers can use them together by feeding outputs from Omniverse simulations into Cosmos models.
Next Steps
Join the Omniverse Community
Connect with Omniverse developers, share feedback, explore samples and learning resources, and stay current on OpenUSD, simulation, and physical AI workflows.
Start Building With Omniverse
Explore accelerated libraries, microservices, and skills for physical AI simulation and agentic workflows.
Omniverse GitHub
Explore Omniverse repositories, samples, and open-source resources for building with OpenUSD, rendering, physics, sensors, and validation.