Cloud infrastructure on your terms

Run apps and AI on your own terms. Your cloud. Your privacy. Your device.

Virtual machines, Kubernetes, storage, databases, networking and AI on servers you can own, down to your own devices. Data stays where you put it, and you pay by the hour with nothing to leave.

See what you can run Join the waitlist

No credit card to start. One small machine, 10 GB of storage and a database are free every month.⁠[1]

  • 1,800+ nodes
  • SOC 2 Type II attested
  • No exit fee

  • $0.024/hr

    for a 2 vCPU Linux machine with 4 GB and local NVMe.⁠[1]

  • 2 to 5 min

    from a click to a running Kubernetes cluster.⁠[2]

  • $0

    to leave. Outbound traffic from $0.01 per GB, inbound free.⁠[1]

  • 1,800+

    nodes in the network your private cloud can fail over to, company-stated.⁠[3]

Prices in U.S. dollars from the published rate card, billed by the hour with no minimum term. The node count comes from the platform operator's own inventory and is not verifiable by a third party.

What makes it different

Six things a public cloud will not give you.

  • Servers you choose and can own

    Your racks, Triple X Cloud nodes, or both. You know where every byte sits.

  • One private network per customer

    A country, an agency or a carrier each runs apart. Nothing is reachable unless you publish it.

  • Placement pinned first

    You choose where data lives before anything is scheduled, and it is not moved to be processed.

  • Sealed: unreadable in use

    For the work no operator may see, data stays unreadable while it runs, and a receipt proves it.

  • Failover across 1,800+ nodes

    If a server fails, work moves across the network under the policy you set.⁠[3]

  • By the hour, nothing to leave

    No minimum term, no exit fee. Standard APIs on every product, so moving out is a copy, not a project.⁠[1]

Products

What you can run.

Open a product for what it is, what it costs and the docs.

  • Virtual machines

    Linux servers with SSH, billed by the hourFrom $0.024 an hour

  • Kubernetes

    Managed clusters, up in 2 to 5 minutesNodes from $0.024 an hour

  • Object storage

    S3-compatible buckets, one endpoint$0.02 per GB a month

  • Managed databases

    PostgreSQL 17, three nodes, 7-day restoreQuoted per size

  • Container registry

    Private images for Docker and your clustersPushes free

  • Networking

    One private network per customerInbound free, no exit fee

  • AI and GPU

    Private inference, GPU on quote, SealedQuoted per GPU-hour

  • Docs, API, CLI and MCP

    REST API, CLI, SDKs, and an MCP server for agentsDocs and MCP server

Who it is for

Whichever term matters most to you.

Some buyers must keep data where it is. Some must keep it private. Some must keep the bill down. The same cloud serves all three.

How it works

One cloud. Three ways in.

Our nodes, your nodes, or any mix of clouds and devices.

  1. 01Triple X Cloud nodesOurs. Start in minutes.Ours
  2. 02Triple X Cloud Nodes you ownBuy A10 or R5 nodes. Own them.Yours
  3. 03Multi-cloud, multi-deviceOurs, yours, your hardware, your clouds.AzureGoogle CloudAny mix

Run by Triple X Cloud

Your private cloudOne private network. Same products.

  • Your customersAt the addresses you publish.
  • Your teamConsole, API and CLI.
  • Your AI agentsYour own inference endpoint.
Three ways in at the top, Triple X Cloud nodes, Triple X Cloud Nodes you own, or a mix of clouds and devices including AWS, Azure and Google Cloud, feed one private cloud run by Triple X Cloud, with virtual machines, Kubernetes, storage, databases, AI and GPU and Sealed inside. Out of it, your apps and AI reach your customers, your team and your AI agents. One hourly bill. Nothing to leave.
What runs underneath, and who operates it

Pricing

Hourly. No minimum term. No exit fee.

Like for like, Tier B runs about 10% below AWS, Azure and Google. Move eligible workloads to Tier D and that reaches 30%. Tier A costs 15% more, and should.⁠[4]

TierA, RegulatedB, EnterpriseC, EdgeD, Public
Price index115908070
Uptime commitment99.99%99.95%99.9%99.5%
FitsRegulated data, residency obligations, public sectorProduction work with no placement constraint. Most workloads land here.Latency-bound work near the point of useRestartable work: batch, CI, rendering, dev and test

Indexed to hyperscaler on-demand list price at 100, modelled against published on-demand rates, July 2026, not measured from customer migrations.⁠[4] The uptime figures are commitments in the master services agreement.

What the hour includes

The machine or node, its memory and local NVMe. Inbound traffic and traffic inside the same region are free. Outbound is $0.01 per GB on the first terabyte, $0.008 to 10 TB and $0.006 beyond.⁠[1]

Free to try

One small virtual machine for 750 hours a month, 10 GB of object storage, 1 GB of outbound transfer and one shared-CPU database. No credit card.⁠[1]

Budgets, alerts and quotas

Set a budget with alert thresholds and email notice; spend is broken down by service, region, resource and period. Starter accounts begin with 5 machines, 2 clusters, 8 cores and 16 GB; increases are reviewed in 1 to 2 business days.⁠[1]

Every price on one page

Get started

Start on the free tier today.

Or talk to Triple X about running it on your own devices.

  1. 01

    Join the waitlist.

    Email and a name. No credit card.

    TakesA minute

  2. 02

    Launch a machine or a cluster.

    From the console, the API or the CLI.

    Takes2 to 5 minutes

  3. 03

    Bring your own servers.

    Triple X sets up your private cloud on them.

    WhoTriple X

We never ask for your root credentials, API keys or bank details by email.

Join the waitlist Talk to Triple X

Questions

Ask by the term that matters.

Every answer ends with where it comes from.

Show questions about

8 answers

All questions
Does my data stay where I put it?

Yes. Your private cloud runs on the servers you chose, and the placement is pinned before anything is scheduled. Failover follows the policy you set.

Proofyour private cloud's placement policy, shown in the console and the API.

Can I use servers I already own?

Yes. Triple X connects your racks to your private cloud, adds Triple X Cloud nodes if you want them, and you manage both from one console.

Proofthe server list in your private cloud names each machine and who owns it.

Who can read my data?

Your private cloud's own people, on its own private network. The platform underneath schedules and bills: it sees that a workload exists and how big it is, not what is in it.

Proofthe platform's SOC 2 Type II report, under NDA, describes operator access and logging.⁠[5]

What is Sealed?

Confidential inference for the data no operator may see. The data stays unreadable while the model runs on it, and a receipt proves the run happened that way.

Proofthe receipt issued for every Sealed run.

What does it cost to leave?

The transfer rate and nothing else: from $0.01 per GB out, with no exit fee, in the contract. Every product uses a standard API, so you leave with a copy, not a migration project.

Proofthe published transfer bands and the no-exit-fee term.⁠[1]

How is the bill worked out?

By the hour, per machine or node, plus outbound transfer. No minimum term. Set a budget with alerts, and read the spend by service, region and resource.

Proofthe rate card and the billing docs.⁠[1]

Is it standard Kubernetes?

Yes. The managed clusters pass 428 of 428 conformance tests; kubectl, Helm and your manifests work as they do anywhere. You can also import a cluster you already run.

Proofthe conformance run, listing with CNCF in progress.⁠[6]

What happens if a server fails?

Work moves to other nodes under the policy you set, inside your private cloud first and across the network of 1,800+ nodes if you allow it.

Proofthe failover policy on your private cloud; the fleet count is company-stated.⁠[3]