Managed Soperator
Nebius-managed Kubernetes operator that runs Slurm clusters on GPU infrastructure for fault-tolerant large-scale AI training, with topology-aware scheduling and automatic node health checks and recovery.
Find alternatives to Managed Soperator
nebius.com/prices itemises 'Managed Soperator (Slurm on Kubernetes)' as Free, footnoted 'Free software, consumption-based pricing'. The product page states the managed service is charged on consumption of the underlying compute only, and that Managed Kubernetes costs $0 when used under Managed Soperator. Underlying GPU compute bills per GPU-hour on the same price list.
What it does
- Model training
- GPU cloud
Sources
- Pricing
-
'Managed Soperator (Slurm on Kubernetes)' listed as Free in the Other services table, independently confirmed on re-fetch.
- Sold within
-
Official documentation · 6 Sep 2026
The managed service has no price of its own and is billed through AI Cloud compute consumption; there is no separate signup or checkout. The open-source Soperator build can be self-hosted cloud-agnostically, but that is the OSS build, not the managed service.
- Name
-
Price-list line item reads 'Managed Soperator (Slurm on Kubernetes)'. The product page distinguishes three delivery models: Managed Soperator (managed by Nebius), Professional Soperator (professional-services install) and Soperator (open source).
- URL
-
Official documentation · 6 Sep 2026
Title 'Slurm-on-Kubernetes Solutions', H1 'Managed Soperator': 'A fully managed Slurm-on-Kubernetes solution for simplified AI training on NVIDIA GPU clusters.' Independently re-fetched.
- Deployment
-
Official documentation · 6 Sep 2026
Managed delivery on Nebius AI Cloud (saas); the open-source build is described as deployable in cloud-agnostic environments (self-hosted).
Also from Nebius Group
-
Nebius AI Cloud Infrastructure service
Rented GPU clusters with storage and networking for training and serving models, billed by the GPU-hour.
-
Nebius Token Factory Model API
Managed inference endpoint serving open-weight models on Nebius hardware, billed per token.
-
Serverless AI Infrastructure service
Nebius AI Cloud's on-demand GPU runtime that runs containerised AI workloads as Jobs and hosts custom models behind HTTP Endpoints without provisioning or managing clusters.
-
Managed Service for MLflow Developer tool
Fully managed MLflow deployment on Nebius AI Cloud for tracking experiments, metrics and artifacts across the machine-learning lifecycle without maintaining tracking-server infrastructure.
Open Nebius Group in the directory
Something wrong here? Send a correction — quote this product id: nebius-managed-soperator.