For teams that build and ship every day

Three build server plans, from daily to quarterly

Every plan gives you an entire dedicated physical Mac mini—not a virtual machine. The data centre handles power, networking, and cooling, with service available 365 days a year so your team can focus on building and shipping.

Choose a plan based on your project's memory and cache needs, then pick a billing period based on how long you need the server to run. Plans start at $19.7/day. If you only build occasionally and already have an idle Mac, self-hosting may still be the right choice.

All prices are listed in USD. Check the portal for real-time server availability.

Start with your budget

Pricing changes with the term; your whole server stays dedicated

All three plans include the macOS graphical interface and command-line tools. Test with a representative project to check memory use, dependency downloads, and build artifact sizes. There’s no need to choose the top plan for concurrency you haven’t needed yet.

Port M4 Base

For a single app and a few builds a day—get one pipeline running reliably.

$19.7/day

Chip
M4
Unified memory
16GB
SSD
256GB
Daily
$19.7/day
Weekly
$53.2/week
Monthly
$98.5/month
Quarterly
$267.9/quarter

A good fit for validating code signing, archiving, and uploads. Avoid running multiple simulators alongside several heavy builds in 16GB of memory.

Rent Port M4 Base

Port M4 Plus

For daily merges, multi-branch builds, and simulator testing, with more room for caches.

$40.3/day

Chip
M4
Unified memory
24GB
SSD
512GB
Daily
$40.3/day
Weekly
$108.9/week
Monthly
$201.6/month
Quarterly
$548.4/quarter

A good fit for always-on runners, reusable dependency caches, and scheduled tests. Set concurrency based on your project's measured workload, not your team size.

Rent Port M4 Plus

Port M4 Pro

For large multi-target projects, heavier test workloads, and local model inference.

$61.4/day

Chip
M4 Pro
Unified memory
64GB
SSD
2TB
Daily
$61.4/day
Weekly
$165.8/week
Monthly
$307.1/month
Quarterly
$835.3/quarter

For workloads where memory or project size is a proven bottleneck. If you need multi-GPU training, a single Mac mini isn’t the right compute setup.

Rent Port M4 Pro

All charges are in USD. Payments are accepted via USDT-TRC20 and Visa / Mastercard / Amex (through Stripe). Available payment gateways are determined by the backend at checkout.

Start with your pipeline bottleneck

Find out what’s causing the queue before upgrading

Waiting for resources on a shared runner is different from tasks waiting on each other within a dedicated server. A dedicated server removes contention with other customers, but you still need to manage concurrency, disk space, and task scheduling on the same machine.

Choose a plan for your workload; these are starting points to test, not fixed concurrency guarantees
Project and workloadSuggested starting planWhat to test firstAvoid
Light builds: one app, a few builds a dayPort M4 BaseCan a full archive complete? Do dependencies, DerivedData, and artifacts fit within your disk budget?16GB isn’t suited to running multiple simulators at once or mixing in heavy inference workloads.
Daily development, parallel branches, and simulator testingPort M4 PlusMemory pressure while builds and tests run together; whether main and nightly jobs need to run at different times.Multiple registered runners don’t mean unlimited parallel builds. Set a limit on simultaneous jobs first.
Large multi-target projects and large-model inferencePort M4 ProModel working set, peak project memory, and archive size; whether builds and inference can run at different times.Not suited to multi-GPU training. If the bottleneck is network access to dependency sources, a faster chip may not help.

Save logs from a clean build and a cache-hit build, then compare total time, memory pressure, and free disk space. If you only build occasionally and already have a suitable office Mac, self-hosting may cost less.

Full hardware details

A dedicated physical server, not a shared allocation

The entire machine is assigned to one customer. You can run multiple jobs for your team, but they still share the server’s CPU, unified memory, and SSD. Scheduling jobs on one machine isn’t the same as running them on separate isolated devices.

Port M4 Base

Chip
Apple Silicon M4
Unified memory
16GB
Storage
256GB SSD
Network
Connected at the selected data centre; 5 regions available
Device type
Dedicated physical machine, not a VM

Port M4 Plus

Chip
Apple Silicon M4
Unified memory
24GB
Storage
512GB SSD
Network
Connected at the selected data centre; 5 regions available
Device type
Dedicated physical machine, not a VM

Port M4 Pro

Chip
Apple Silicon M4 Pro
Unified memory
64GB
Storage
2TB SSD
Network
Connected at the selected data centre; 5 regions available
Device type
Dedicated physical machine, not a VM

Regions: Singapore, Japan (Tokyo), South Korea (Seoul), Hong Kong, and US West. Specs don’t guarantee a particular build speed; project dependencies, Xcode version, and caching strategy all affect results.

Add only what you need

More cache space and multi-machine connections are separate needs

First work out whether you’re short on disk space or memory. An SSD upgrade gives you more room for dependencies, archives, and caches—it doesn’t add unified memory. Thunderbolt 5 linking is for coordinating multiple machines; it won’t combine them into one machine with more memory.

+1TB SSD

For projects approaching their base storage limit that don’t yet need a larger upgrade. Clear rebuildable old caches first, then decide whether you need to keep more archived builds.

Daily
$2.6/day
Weekly
$7/week
Monthly
$12.9/month
Quarterly
$35.1/quarter

+2TB SSD

For large build artifacts, batch exports, or longer archive retention. Extra capacity isn’t a substitute for a separate backup—store important certificates and deliverables elsewhere too.

Daily
$5.2/day
Weekly
$14/week
Monthly
$25.8/month
Quarterly
$70.2/quarter

Thunderbolt 5 linking

Priced per machine, for setups that already need to coordinate multiple machines. Check how jobs will be split and whether your setup supports the connection. Regular iOS builds on one server don’t need this add-on by default.

Daily / machine
$1.4/day
Weekly / machine
$3.7/week
Monthly / machine
$6.9/month
Quarterly / machine
$18.8/quarter

These prices are for add-ons only and don’t include the server rental. Check the total for your selected term, quantity, and configuration at checkout. For multi-machine deployments, confirm your setup requirements through a support ticket in the portal.

Know what makes up your bill

A sample monthly bill, item by item

Here’s a pricing example for a Port M4 Plus in Tokyo with a +1TB SSD add-on. This is a configuration estimate, not a payment receipt or confirmation that a server has been created.

Sample monthly configuration

Billing period
Monthly
Server region
Tokyo
Port M4 Plus
$201.6/month
+1TB SSD
$12.9/month
Total
$214.5 USD
Calculation: 201.6 + 12.9 = 214.5. Your actual order total is shown at checkout.

The billing period sets the price column

This example uses monthly billing, so both the server and add-on are charged at their monthly rates. The monthly price isn’t calculated by multiplying the daily rate by the number of days.

The region shows where the server is deployed

Tokyo is the server location in this example, not an extra regional charge. When choosing a region, consider the location of your code repository, dependency sources, and artifact uploads.

Server and add-on costs are listed separately

The server includes an Apple Silicon M4, 24GB of unified memory, and a 512GB SSD. The +1TB SSD is a separate storage add-on and doesn’t change the chip or memory.

Choose a payment method at checkout

Pay in USD with USDT-TRC20 or Visa / Mastercard / Amex (through Stripe). Available payment gateways are determined by the backend.

Choose based on how long you’ll run it

Use a short term to test; compare monthly pricing for ongoing use

Daily billing isn’t necessarily cheaper for long-term use. Estimate how long you’ll need the server, compare the total prices for the same plan, and choose a new term after your trial if needed.

Daily: test a complete pipeline

Useful for a first setup, troubleshooting your build environment, or adding capacity during a release week. Don’t stop at a successful compile—test signing, archiving, uploads, and log collection too.

Save reproducible installation and configuration steps so the trial doesn’t leave you with a working directory you can’t recreate.

Weekly: cover a focused work period

Useful for time-limited integration, testing, and release work. If the work spans multiple weeks, compare the monthly price directly rather than assuming weekly billing is still the best option.

Choose a term that covers time spent waiting and reviewing, not just the estimated hours of compilation.

Monthly or quarterly: keep a persistent environment

For teams with regular merges, nightly tests, and reusable caches. Before choosing a quarterly term, make sure the project will keep running and the plan can handle the workload ahead.

A longer term means configuring the environment less often, but you’re still responsible for backups and pipeline configuration. Don’t keep the only copy on the server.

A quick way to compare: Divide the monthly price for your plan by its daily price to find the number of daily-billed days that would cost about the same as a month. Base: 98.5 ÷ 19.7 = 5; Plus: 201.6 ÷ 40.3 ≈ 5.003; Pro: 307.1 ÷ 61.4 ≈ 5.002. If you need a server for 6 consecutive days, compare monthly billing for all three plans first. This is a price comparison only; existing orders won’t automatically switch billing periods.

One catalogue, five regions

Choose a region based on your pipeline’s network path

All three plans are available in Singapore, Japan (Tokyo), South Korea (Seoul), Hong Kong, and US West. Proximity isn’t just about distance from your team; consider the route to your code repository, dependency sources, and artifact destinations. We don’t make unmeasured latency claims for individual cities.

All listed plan and region combinations have sufficient availability. Check the portal for real-time status; this table isn’t a live availability check.
PlanSingaporeJapan (Tokyo)South Korea (Seoul)Hong KongUS West
Port M4 BaseAvailableAvailableAvailableAvailableAvailable
Port M4 PlusAvailableAvailableAvailableAvailableAvailable
Port M4 ProAvailableAvailableAvailableAvailableAvailable

During your trial, record the time for code checkout, the first dependency install, a cached build, and artifact upload. If external dependency downloads are slow, try adjusting your caching and network path before upgrading your plan.

Know the details before you order

You can change plans; prepare your data in advance

Check the rental term, server changes, and payment details before relying on a server for daily builds. For actions on an existing order, follow the portal and confirm through a support ticket.

Can I upgrade if I run out of memory?

Submit a support ticket through the portal with your current plan, peak project memory use, and requested configuration. Changing physical hardware may require a server replacement and environment migration; it isn’t like adding memory to a VM online. Confirm the available option, cost, and migration plan before proceeding, and back up your code, configuration, signing materials, and important artifacts first.

What should my team check before renewal or expiry?

Check the order term, expiry time, and renewal options in the portal. Don’t assume one payment automatically extends every server. Renew or move your work before expiry, and check nightly jobs, certificate backups, and artifact storage in particular. A rented server isn’t a long-term archive you can count on accessing after expiry. The portal shows the current status of your order.

Why are plans and add-ons priced only in USD?

Using one currency makes it easier to compare servers, rental terms, and add-ons on the same price list. Prices aren’t converted to local currencies on this page. Before paying, check the USD total at checkout and confirm the selected configuration and billing period so you don’t compare a daily rate with a monthly one.

Which payment methods can I use?

We accept USDT-TRC20 and Visa / Mastercard / Amex (through Stripe) only. Available payment gateways are determined by the backend. If you need help checking an order after payment, include your order details and relevant payment records in a support ticket. Never send your full card number or other sensitive credentials.

Does dedicated access mean builds never have to wait?

Dedicated access means the entire physical machine is assigned to one customer, so you don’t compete with other customers for its resources. Builds on the same server still share CPU, memory, and disk. Set a concurrency limit for your runners and schedule heavy UI tests and archives at different times. Decide whether you need multiple servers based on how often jobs arrive and how long they take to complete.

Does year-round availability mean I don’t need backups?

No. Servers are available 365 days a year, but hardware issues may still need troubleshooting and replacement through a support ticket and the confirmed process. Back up signing materials, pipeline configuration, and artifacts that can’t be recreated elsewhere. Caches can be rebuilt, but allow for the recovery time when planning releases. Keep the steps to rebuild your environment so you can reduce manual work if a server needs to be replaced.

Start with a day and run your full pipeline

Plans start at $19.7/day. Test builds, tests, archiving, and uploads with your own repository, then decide whether to keep the server running. There’s no need to commit to a long term before you’ve tested your actual workload.