Pricing
From DIY to fully managed
clusterdOS
The open-source OS engine that runs inside a cluster.
Free
/self-hosted
Open source. Run it on your own hardware.
Get clusterdOS
Features
Operates below the workload layer, on the hardware itself
Discovers, diagnoses and resolves GPU thermal events, ECC errors, networking faults
Distributed OS engine built on Kubernetes
Extends to VM-based and Slurm environments
Cilium/eBPF network layer
NVIDIA and AMD driver stacks
Support
Community Slack
Aranya multicluster OS
The multicluster operating system that orchestrates across your fleet. You run the software.
Coming soon
Pricing not yet announced.
Join the waitlist
Features
Federated architecture: reason about every cluster as one system
Direct your fleet in plain language
Role-based permissions at the OS level. Access is granted, not assumed
Observability from one GPU temperature to federation-wide traffic
Automates the operational work a platform team would do by hand
Recover cost by recycling idle compute across the fleet
Support
Shared Slack channel
Managed observability dashboards
Fully managed clusters
We design the architecture, stand it up, and run it end-to-end. Bare metal to production in 48 hours.
$0.06 – $0.16
/GPU/hr
pricing based on requirements.
Talk to an engineer
What's included
Aranya Multicluster OS, fully managed
Production-ready cluster in 48 hours
Custom architecture: Kubernetes, VMs, or Slurm
Custom Networking architecture
24/7 on-call and incident response
Capacity scaling and ongoing platform engineering
Utilization monitoring to rightsize your deployment
managed storage (add-on)
RECOMMENDED
Support
Shared Slack channel
Managed observability dashboards
Aranya automates all the moving parts from individual kernel drivers to federation across clusters at unprecedented scale.
