Deploy an agenticworkforce environment.

Any harness. Any model. Any cloud. Run it on ours and start in a minute, or run it on yours with our engineers alongside you.

01 · XO Cloud (Managed)

Pick a harness, pick a model, and be running in a minute.

We host it and we run it. There is no licence to negotiate and no call to book: sign up, choose a harness, and the environment is live on our cloud. Every one reports to the same unit-of-work ledger, so spend is legible from the first run.

30 days free. Priced per environment, never per seat.

Included with every harness

  • Command via Slack · WhatsApp · Telegram · Discord
  • Switch seamlessly across 100+ AI providers
  • Connect to VS Code · Cursor · GitHub · Vercel

Choose your harness

7 harnesses · scroll for more

02 · Bring your own cloud

Your cloud, your hardware, stood up by our engineers.

The same product, licensed onto infrastructure you already own. You connect the clouds, accelerators, models and harnesses your team runs today, and our forward-deployed engineers install it with you and shape it around how your business actually works.

This is the lane for larger teams: work that is not allowed to leave a boundary, an air-gapped network, a regulator who needs the ledger on your side, or a team that has to put its own brand in front of the people using it.

Connect what you already run

  • Google Cloud
  • Azure
  • Nebius
  • Nvidia
  • AMD
  • Groq

Anywhere else? Tell us where it needs to run.

What the licence hands over

01

Your estate

It runs in your cloud, on your own hardware, or fully air-gapped. Nothing has to cross your boundary for the product to work.

02

Your brand

White-labelled throughout. The console your team opens is yours, not ours, and the people using it need never learn our name.

03

Your numbers

Unit-of-work metering runs on your side and exports. The ledger of what the agents delivered and what it cost is yours to keep and to audit.

04

Your release train

Pinned, supported releases rather than automatic updates, so a version change is a decision your team makes on its own schedule.

How it lands

  1. 01

    Scope

    Tell us where it needs to run: which cloud or hardware, which accelerators, and which compliance regime it has to sit inside.

  2. 02

    Stand it up

    Our forward-deployed engineers install it on your infrastructure alongside your team, connecting the models and harnesses you already run. No rip-and-replace.

  3. 03

    Customise

    The machine speed engagement: blueprints adapted to your operation and pointed at the metric that matters this quarter, built on the stack you already own. See machine speed.

  4. 04

    Run it

    A named contact and a response SLA, with SSO and SAML, audit trails and data residency configured and managed.

Try it on your own machine first

The runtime is open source. One command puts the same thing that runs under every managed environment onto your hardware, with the full observability layer attached.

curl -fsSL quirq.ai/install | sh

Runs anywhere · macOS, Linux, your cloud

Pricing on application · deployment and customisation included

Talk to us

Either way

The same product. Wherever it runs.

Managed or licensed, an environment is the same three layers with the same record underneath. Only the address changes.

01

Spaces

The ground floor. One space holds many environments, across any cloud, any hardware, or an air-gapped network.

02

Environments

Where the work happens. Any harness, any model, with the record of everything that ran inside.

03

Observatory

The view over both. Graph, timeline and sessions: what ran, what it cost, what data it touched.

Three readings of the same record

01

Graph

Every environment, artifact and cross-tie in the space: clustered, searchable, live.

02

Timeline

Scrub the whole history: how each environment grew, month by month, run by run.

03

Sessions

Every input, tool call, model and token: what was spent, and what data it touched.

Side by side

Two lanes. One decision.

The rows that describe what the product does read the same in both columns: every harness, every model, and the full observability layer. What changes is where it runs, who stands it up, and whose name is on it.

XO Cloud (Managed) and bring your own cloud, compared.
CapabilityXO Cloud (Managed)Hosted by us · self-serveBring your own cloudYour estate · licensed
How you startSign up and goTalk to us
Where it runsOur cloudYour cloud, on-prem or air-gapped
Who stands it upYou, in a clickOur engineers, with your team
Environments30 · 500 by planUnlimited
All harnesses and models yes yes
Observability: graph, timeline, sessions yes yes
Unit-of-work meteringOn our infraYour numbers, exportable
One-click deploy UI yes yes
Custom workflows and blueprintsnot included yes
White-labelnot included yes
SSO / SAML · audit · residencyOn Max✓ managed
UpdatesAutomaticPinned, supported releases
SupportEmail · Slack on MaxNamed contact + SLA

Pricing

Pay per environment. Never per seat.

XO Cloud (Managed) is self-serve: pick a tier and start. Bringing your own cloud is quoted, because what it costs depends on where it has to run and how much of it we build with you.

Most popular

Pro

XO Cloud (Managed)

$20/mo

  • 30 managed environments
  • One-click deploy, hosted by us
  • Custom templates and MCP integrations
  • Spend tracking and agent comparison
Start free on the Pro plan (opens in a new tab)

30 days free · no licence needed

Licensed

Bring your own cloud

Custom

  • The managed product on your own infrastructure
  • Stood up and customised by our engineers
  • White-label: your brand throughout
  • Named contact, response SLA, pinned releases

Bringing your own cloud? Tell us where it has to run.

Ask for the deployment guide and a walkthrough of the unit-of-work ledger with your own numbers. We reply within one business day. If you are a smaller team, you do not need us at all: start on XO Cloud (Managed) and skip the conversation.