I keep getting the same question from customers and partners: what does it cost to take an Omniverse project into production? Some are still planning around a required software subscription. That requirement changed in May 2026.
NVIDIA’s licensing page, checked again on September 10, says development and production use no longer require an NVIDIA AI Enterprise subscription. Redistribution is also permitted under the applicable terms. The July 1 announcement confirms the change.
Deployment permission and support are separate choices
When someone asks me about production cost, I now separate the subscription question from the support question. A team can use Omniverse without buying enterprise support. That leaves it with community support through the Developer Forums and Discord. An NVIDIA AI Enterprise subscription remains available for teams that need enterprise support.
I would make that choice from the workload: who will diagnose a failure, how long can the service be unavailable, and what support commitments has the team made to its users? An internal experiment and a customer-facing service may need different answers.
The license still applies
The NVIDIA Software License Agreement and product-specific terms still govern use. Removing a required subscription does not remove those terms, or the hardware and operating costs of running the application.
The separate move toward standalone libraries changes how I build on Omniverse. This licensing change answers a narrower question: whether production use itself requires the subscription. It does not tell me whether an early-access library is ready for a particular deployment.
If a project was put aside because of the old subscription requirement, I would revisit that assumption. Then check the terms for what is being shipped and decide who will support it.
Disclaimer: The views and opinions expressed in this account are those of my own and do not represent those of my employer, NVIDIA.