Product

Solutions

Pricing

Company

See Orion in action

Product

Solutions

Pricing

Company

See Orion in action

Products

Pricing

Solutions

Integrations

Company

See Orion in action

One compute plane. Every workload.

Orion orchestrates GPU workloads, containers, VMs, and bare metal from one unified compute plane — so your team manages one system instead of five.

See Orion in action

Read the docs

The fragmentation problem.

Most teams manage separate systems for containers, VMs, and bare metal. That means three ops workflows, three billing systems, and three sets of failure points.

VMware costs are up 150% to 10×+

vSphere 7 is already out of support. vSphere 8 end of general support: October 2027. The migration window is narrowing — and every path forward involves rebuilding your infrastructure.

GPU hardware sits idle most of the time

Enterprise on-premises GPU utilization sits at 10–15%. The hardware is paid for. The capacity is there. Most environments lack the orchestration layer to actually use it.

Three clusters for three substrate types

Kubernetes for containers. VMware for VMs. Custom tooling for bare metal. Each has its own admin workflow, billing model, and failure mode. Your team manages the seams, not the work.

Provisioning measured in days, not minutes

Getting a researcher or artist a GPU workstation means a ticket, a queue, and someone from ops in the loop. That friction adds up. Orion provisions in 60 seconds, without IT involvement.

GPU, VM, container, bare metal. One cluster.

GPU Operator Automation

Typically 2–4× more workload density

Orion automates NVIDIA, AMD, and Intel GPU operator installation and configuration. Admins choose their slicing method (MIG, vGPU, or time slicing) through a UI. No YAML. No manual node labeling. End users get more capacity without knowing it exists.

GPU Operator Automation

Typically 2–4× more workload density

GPU Operator Automation

Typically 2–4× more workload density

Autoscaling

Right-node, right-size autoscaling

When demand spikes, Orion selects the right node for the job — not just the largest available. Scale up on burst, scale back when idle. No wasted capacity, no over-provisioning.

Autoscaling

Right-node, right-size autoscaling

When demand spikes, Orion selects the right node for the job — not just the largest available. Scale up on burst, scale back when idle. No wasted capacity, no over-provisioning.

Autoscaling

Right-node, right-size autoscaling

When demand spikes, Orion selects the right node for the job — not just the largest available. Scale up on burst, scale back when idle. No wasted capacity, no over-provisioning.

Load Balancing

Request-aware load balancing

Orion distributes workloads evenly across your fleet by active request count, not just CPU/memory headroom. Your cluster stays balanced without manual intervention.

Load Balancing

Request-aware load balancing

Orion distributes workloads evenly across your fleet by active request count, not just CPU/memory headroom. Your cluster stays balanced without manual intervention.

Load Balancing

Request-aware load balancing

Orion distributes workloads evenly across your fleet by active request count, not just CPU/memory headroom. Your cluster stays balanced without manual intervention.

Provisioning

60-Second Provisioning

No tickets. No JIRA queues. No waiting for IT. Orion provisions containerized and virtualized workloads on demand — researchers and artists get their environment before they've finished their coffee.

Provisioning

60-Second Provisioning

Provisioning

60-Second Provisioning

Multi-Cloud Orchestration

Crossplane: no drift, no lock-in.

Via Crossplane (available as a Terra plugin), Orion enforces your infrastructure configuration continuously — unlike Terraform or Pulumi, which apply once and drift. Mount AWS, GCP, Azure, or on-prem as a single abstraction. Deploy the same way everywhere. If a cloud raises prices or goes down, you have a path out.

Multi-Cloud Orchestration

Crossplane: no drift, no lock-in.

Your entire fleet, one view.

Orion sits between your infrastructure and your workloads — scheduling containers, VMs, and bare metal jobs across GPU and CPU resources with unified visibility across the whole fleet.

Unified Compute Plane · Deploy on-premises, in the cloud, or in hybrid and air-gapped configurations

API / kubectl / GitOps

Admin

End User

Web-based access · Projects & Workloads · User & Group Management · Storage · Networking · License Management

Observability & Metrics

OpenTelemetry

Prometheus · Grafana

Node visibility: service, workstation, headless

Service health topology

Multi-tenant namespace isolation

Chargeback · Cost attribution

API token management

App Store

Open-source Kubernetes app marketplace

Helm chart-based plugin system

One-click deployment via ArgoCD

Git-sourced plugin repositories

Official Juno + community plugins

Custom private source repositories

Namespace-scoped install management

Primary user access point · Launch & connect to workloads · Session sharing · Role-based views · Service topology

GPU-Aware Scheduling

Fractional GPU · Time-slicing

Bin-packing · NVIDIA MIG / vGPU

WebRTC containerized desktop streaming

Containers · VMs via KubeVirt

Windows GPU passthrough

Ephemeral on-demand environments

Share & connect to running workloads

Workstation Delivery

Browser-based workstation delivery

Web access · Collaboration

GitOps managed workload definitions

Project-scoped resource access

Session sharing · Workload connect

Zero Kubernetes knowledge required

Quickbar · Service topology widget

Rhea auth/authz on all service calls

Orchestrates

Infrastructure

On-Prem

Full air-gap support

On-Prem Data Center

Air-Gapped ★

Hybrid

On-prem + cloud bursting

On-Prem

AWS / Azure Cloud Burst

Multi-Cloud

Any cloud, any combination

AWS

Azure

GCP

CoreWeave

Oracle

Integrates With

Kubernetes

EKS · AKS · GKE · OpenShift · RKE2 · K3s

Identity

Okta · Azure AD · Google · AWS Cognito · OIDC

Storage

NFS · S3 · Qumulo · Weka · Vast · Any CSI

Observability

Prometheus · Grafana · OpenTelemetry

Compute

NVIDIA · AMD · Intel GPUs · KubeVirt VMs

Key Differentiators

✓ Zero CRDs — native Kubernetes primitives only · ✓ 1–2 minute OneClick install · ✓ Air-gapped / classified deployment built-in · ✓ No Kubernetes knowledge required for end users · ✓ Any workload: containers · VMs · GPUs · bare metal

Unified Compute Plane · Deploy on-premises, in the cloud, or in hybrid and air-gapped configurations

API / kubectl / GitOps

Admin

End User

Web-based access · Projects & Workloads · User & Group Management · Storage · Networking · License Management

Observability & Metrics

OpenTelemetry

Prometheus · Grafana

Node visibility: service, workstation, headless

Service health topology

Multi-tenant namespace isolation

Chargeback · Cost attribution

API token management

App Store

Open-source Kubernetes app marketplace

Helm chart-based plugin system

One-click deployment via ArgoCD

Git-sourced plugin repositories

Official Juno + community plugins

Custom private source repositories

Namespace-scoped install management

Primary user access point · Launch & connect to workloads · Session sharing · Role-based views · Service topology

GPU-Aware Scheduling

Fractional GPU · Time-slicing

Bin-packing · NVIDIA MIG / vGPU

WebRTC containerized desktop streaming

Containers · VMs via KubeVirt

Windows GPU passthrough

Ephemeral on-demand environments

Share & connect to running workloads

Workstation Delivery

Browser-based workstation delivery

Web access · Collaboration

GitOps managed workload definitions

Project-scoped resource access

Session sharing · Workload connect

Zero Kubernetes knowledge required

Quickbar · Service topology widget

Rhea auth/authz on all service calls

Orchestrates

Infrastructure

On-Prem

Full air-gap support

On-Prem Data Center

Air-Gapped ★

Hybrid

On-prem + cloud bursting

On-Prem

AWS / Azure Cloud Burst

Multi-Cloud

Any cloud, any combination

AWS

Azure

GCP

CoreWeave

Oracle

Integrates With

Kubernetes

EKS · AKS · GKE · OpenShift · RKE2 · K3s

Identity

Okta · Azure AD · Google · AWS Cognito · OIDC

Storage

NFS · S3 · Qumulo · Weka · Vast · Any CSI

Observability

Prometheus · Grafana · OpenTelemetry

Compute

NVIDIA · AMD · Intel GPUs · KubeVirt VMs

Key Differentiators

The orchestration layer handles scheduling, resource allocation, and lifecycle management across every substrate — so your team doesn't have to.

Customer-hosted

Your data never leaves your perimeter.

Orion deploys in your environment — on-prem, air-gapped, or hybrid. There is no cloud management plane calling home, no vendor access to your cluster, and no egress fees. For life sciences, defense, and enterprise teams where data sovereignty is non-negotiable, this is the architecture that makes Orion viable where others aren't.

No cloud management plane

Orion runs entirely within your network. No AWS account required, no Azure backbone, no external orchestration layer. Your cluster operates independently of any vendor's cloud.

Zero vendor telemetry

No phone-home. No licensing server that needs internet access. Licensing, updates, and orchestration all operate inside your perimeter. Fully air-gapped deployments are production-supported.

No egress surprises

Data stays where you put it. No egress fees, no cross-region transfer, no hidden bandwidth costs. R3D cut their AWS compute bill ~40% in part because their data stopped moving.

Terra App Store

The infrastructure app store. Powered by GitOps.

Once engineers start using Terra, Orion stops being infrastructure — it becomes the developer experience. Terra is Orion's infrastructure app store. Three plugin types cover everything your team needs.

Operators install GPU drivers, AI runtimes, and cluster tooling with opinionated defaults that cover 99% of deployments.

Template Engines define full environments — Helios containerized desktops, JupyterLab, VS Code Server, custom pipelines — delivered to any user with one click.

Network and Services plugins drop in Tailscale exit nodes, NFS provisioners, and connection brokers without touching cluster config. Fork any chart if you need to go deeper.

Terra App Store

The infrastructure app store. Powered by GitOps.

Terra is Orion's infrastructure app store. Operator plugins install GPU drivers, AI runtimes, and cluster tooling with opinionated defaults that cover the vast majority of deployments. Template Engines define full environments — Helios desktops, JupyterLab, VS Code Server, custom pipelines — delivered to any user with one click. Network and Services plugins drop in Tailscale nodes, NFS provisioners, and connection brokers without touching cluster config. Fork any chart to go deeper.

Alex Hatfield

CEO & Co-Founder, Juno Innovations

"The idea was always Lego bricks. You pick the tools your team needs — GPU operators, runtimes, workload templates — click to install, and they just work. All the hard stuff stays on our side. You just build."

Terra App Store

One-click app installs

VS Code Server, JupyterLab, DCC tools, custom pipelines — deploy production-ready environments in seconds, not hours.

Terra App Store

One-click app installs

VS Code Server, JupyterLab, DCC tools, custom pipelines — deploy production-ready environments in seconds, not hours.

Terra App Store

One-click app installs

VS Code Server, JupyterLab, DCC tools, custom pipelines — deploy production-ready environments in seconds, not hours.

Templating

Reusable Templates

Build once, deploy everywhere. Create golden-image templates for dev, staging, and production workflows that work identically across your whole team.

Templating

Reusable Templates

Build once, deploy everywhere. Create golden-image templates for dev, staging, and production workflows that work identically across your whole team.

Templating

Reusable Templates

Build once, deploy everywhere. Create golden-image templates for dev, staging, and production workflows that work identically across your whole team.

Versioning

Full Version History

Track every environment change. Roll back to any previous configuration instantly — no more 'it worked last week' debugging sessions.

Versioning

Full Version History

Track every environment change. Roll back to any previous configuration instantly — no more 'it worked last week' debugging sessions.

Versioning

Full Version History

Track every environment change. Roll back to any previous configuration instantly — no more 'it worked last week' debugging sessions.

Helios Workstations

A workstation for every user. Launched in 60 seconds.

Containerized desktop environments provisioned on demand. Users request the resources they need — GPU, RAM, applications — and Helios delivers a full workstation in under 60 seconds. When the session ends, resources return to the pool. No idle machines. No assigned hardware. No IT queue.

Full desktop, zero footprint

Every Helios workstation is a containerized environment with full GPU access, persistent storage, and the tools your team already uses. VS Code, JupyterLab, Nuke, Houdini, Blender — launched from a browser, destroyed on logout.

Full desktop, zero footprint

Capacity that comes back

Traditional VDI pre-provisions fixed VMs that sit idle 70% of the time. Helios provisions on demand and releases resources when sessions end. R3D Studios doubled artist capacity on the same GPU hardware with this model.

Capacity that comes back

Browser-based, anywhere

Artists and researchers connect through a browser — no VPN client, no fat installer, no IT ticket. Kasm delivers color-accurate streaming with sub-frame latency. Works from the office, from home, or from a hotel lobby.

Browser-based, anywhere

KubeVirt

Windows and Linux VMs, orchestrated like containers.

Orion manages Windows and Linux VMs through KubeVirt — on the same cluster as your containers, with no separate hypervisor stack required.

Orion orchestrates Windows and Linux VMs through KubeVirt on the same Kubernetes cluster as your containers. Windows Server 2019 and 2022, GPU pass-through and vGPU slicing, live migration between nodes without downtime — all managed from the same compute plane. No separate hypervisor stack. No infrastructure consolidation project. Run Adobe Creative Suite on Windows while rendering on Linux, and deploy both with the same 60-second provisioning your containerized workloads get.

Read the docs

Windows app support

Adobe Creative Suite, Autodesk, DaVinci Resolve, and more.

Windows app support

Adobe Creative Suite, Autodesk, DaVinci Resolve, and more.

Windows app support

Adobe Creative Suite, Autodesk, DaVinci Resolve, and more.

GPU pass-through and vGPU slicing

Run GPU-accelerated Windows workloads via KubeVirt.

GPU pass-through and vGPU slicing

Run GPU-accelerated Windows workloads via KubeVirt.

Helm-compatible by default

Import existing Kubernetes workload definitions, templatize them, and make them requestable by end users. No rewriting required.

Helm-compatible by default

Import existing Kubernetes workload definitions, templatize them, and make them requestable by end users. No rewriting required.

Helm-compatible by default

Import existing Kubernetes workload definitions, templatize them, and make them requestable by end users. No rewriting required.

See everything. Operate with confidence.

Orion surfaces utilization, CPU saturation, memory pressure, job throughput, and cost-per-workload across your entire fleet — in real time. Provision in minutes with reusable templates, enforce resource quotas across teams, and run on any storage layer you already own. See where you're getting value and where capacity is sitting idle.

Up and running fast

Provision GPU capacity in minutes with sane defaults and reusable templates. No spreadsheet archaeology, no tribal knowledge required.

Up and running fast

Provision GPU capacity in minutes with sane defaults and reusable templates. No spreadsheet archaeology, no tribal knowledge required.

Built for shared teams

Researchers, engineers, and platform teams all in one place — with resource quotas, queue management, and RBAC so nobody steps on each other.

Built for shared teams

Researchers, engineers, and platform teams all in one place — with resource quotas, queue management, and RBAC so nobody steps on each other.

No lock-in. Your infrastructure, your choice.

Orion connects through standard Kubernetes primitives. No proprietary plugins, no forced migration, no rearchitecture required. Deploy on what you already have.

No lock-in. Your infrastructure, your choice.

Orion connects through standard Kubernetes primitives. No proprietary plugins, no forced migration, no rearchitecture required. Deploy on what you already have.

Works with the storage you already have

NFS, S3, Qumulo, Weka, Vast, and any CSI-compatible provider connect out of the box. Orion plugs into your existing storage layer — no migration, no rearchitecture, no new storage vendor required.

Why infrastructure teams choose Orion

Three capabilities no funded competitor offers simultaneously. Here's how the platforms compare.

Traditional infrastructure management

Manual GPU provisioning — hours of wait time per deployment.

10-15% average GPU utilization — paying for capacity you never use.

Siloed clusters with no unified view across your fleet.

Kubernetes complexity that requires a dedicated platform team.

VMs and containers managed by completely separate tools.

Vendor lock-in with proprietary orchestration layers.

Juno

Purpose-built for compute orchestration — containers, VMs, GPUs, and bare metal.

Typically 2–4× workload density via native GPU operator time slicing — no new hardware.

One compute plane — containers, VMs, and bare metal unified.

Fast provisioning via the Orion dashboard — no complex setup required.

Production-ready in days, not months. Kubernetes-native.

Open standards — no vendor lock-in, ever.

Traditional infrastructure management

Manual GPU provisioning — hours of wait time per deployment.

10-15% average GPU utilization — paying for capacity you never use.

Siloed clusters with no unified view across your fleet.

Kubernetes complexity that requires a dedicated platform team.

VMs and containers managed by completely separate tools.

Vendor lock-in with proprietary orchestration layers.

Purpose-built for compute orchestration — containers, VMs, GPUs, and bare metal.

Typically 2–4× workload density via native GPU operator time slicing — no new hardware.

One compute plane — containers, VMs, and bare metal unified.

Fast provisioning via the Orion dashboard — no complex setup required.

Production-ready in days, not months. Kubernetes-native.

Open standards — no vendor lock-in, ever.

Traditional infrastructure management

Manual GPU provisioning — hours of wait time per deployment.

10-15% average GPU utilization — paying for capacity you never use.

Siloed clusters with no unified view across your fleet.

Kubernetes complexity that requires a dedicated platform team.

VMs and containers managed by completely separate tools.

Vendor lock-in with proprietary orchestration layers.

Purpose-built for compute orchestration — containers, VMs, GPUs, and bare metal.

Typically 2–4× workload density via native GPU operator time slicing — no new hardware.

One compute plane — containers, VMs, and bare metal unified.

Fast provisioning via the Orion dashboard — no complex setup required.

Production-ready in days, not months. Kubernetes-native.

Open standards — no vendor lock-in, ever.

What a working studio says about Orion

Donald Strubler

Head of Technology, R3D Studios

"Orion shifted our focus from finding stability to using the stability to iterate."

~40%

Compute cost reduction — R3D Studios

60 sec

User request to workload running — R3D Studios

2:1

GPU density — same hardware, more artists

Frequently asked questions

What is Orion and how does it work?

Orion is a containerized workload platform for on-prem, cloud, or hybrid deployments with GPU time-slicing and auto-scaling.

Can I use my existing pipelines and toolchains?

Yes, Orion mounts your existing file shares and toolchain locations into containers

How fast can new workstations be provisioned?

2-5 seconds with cached images, 1-3 minutes for new nodes depending on workload type.

Can Orion run on older hardware?

Yes, from Raspberry Pis to enterprise-grade - anything that can run containers

Do I need a GRID license for GPU time-slicing?

No license needed for default time-slicing. GRID/vGPU licensing only required for MIG mode.

What cloud providers do you support?

Any Kubernetes cluster - EKS, AKS, GKE, or on-premises. We're 100% cloud-agnostic.

How does licensing work?

Per user, per month with volume discounts for larger teams. We will be moving to a node and consumption based model this year.

See Orion in your environment.

Most teams are getting 10–15% GPU utilization out of hardware they've already paid for. Orion changes that — without a rip-and-replace. Talk to us about your workload profile.

Request a demo →