NVIDIA Open-Sources the Full Run:ai Platform: GPU Orchestration Without a Paid Subscription

What happened

On October 13, 2026, NVIDIA announced the complete Run:ai platform as open source software within its DSX AI Factory ecosystem. According to the company's developer blog, customers, developers and ecosystem partners no longer need a paid subscription or enterprise support entitlement to use its production-grade AI workload and GPU orchestration.

The release covers the integrated platform plus four modules that can be adopted independently: KAI Scheduler, GPU Fractioning Operator, KAI Resource Management and Workload Map. KAI Scheduler is open under the Apache 2.0 license and is a Cloud Native Computing Foundation Sandbox project; NVIDIA says all of its further development happens in the community repository.

Background: NVIDIA completed its acquisition of Run:ai on December 30, 2024 (reported by TechCrunch at the time). This release moves the platform from proprietary distribution to open source. The community edition carries no enterprise support; organizations that need support can use NVIDIA's commercial offerings via channels such as NVIDIA AI Enterprise.

On the ecosystem side, eight partners including Crusoe, Nebius, Mirantis and Rafay announced integrations in parallel. The platform source lives in the DSX AI Factory GitHub organization and is also distributed as a Community Product on NVIDIA NGC, installable via a Helm chart.

Key facts

1. What shipped: The complete Run:ai platform is open source within the DSX AI Factory ecosystem; no paid subscription or support entitlement required.
2. Four modules: KAI Scheduler, GPU Fractioning Operator, KAI Resource Management and Workload Map can be adopted independently; KAI Scheduler is Apache 2.0 and a CNCF Sandbox project.
3. Background: NVIDIA completed its acquisition of Run:ai on December 30, 2024 — nearly two years ago.
4. Ecosystem and editions: Eight partners including Crusoe, Nebius, Mirantis and Rafay announced integrations; the community edition has no enterprise support, while commercial editions remain available via NVIDIA AI Enterprise and other channels.