A whole Mac mini for your build queue

2 Models, 4 Rental Terms

Dedicated physical Macs, not VMs. All 6 locations are available, from $20.1/day. Check your workload’s memory and storage needs, then choose a rental term.

Rent a Cloud Mac Compare Configurations

All prices are in USD. Live availability, order totals, and delivery details are shown in the console.

1 physical machine, dedicated resources Daily / weekly / monthly / quarterly billing SSH and macOS desktop access
Choose the right build capacity

One app archive or multiple parallel jobs?

Both plans include a whole Mac mini, with no CPU, memory, or storage shared with other customers. The difference is headroom: M4 works well for dedicated routine builds, while M4 Pro is suited to parallel testing, larger projects, and local inference.

Routine builds & dedicated CI

OSXBuild Bake M4

Run one app’s archive, signing, and export on a dedicated machine. Keep dependency caches on your own machine instead of setting up the environment from scratch for every build.

Chip
M4
Memory
16GB
SSD
256GB
$20.1 / day

Daily price shown. See the pricing below for other rental terms.

Daily
$20.1 / day
Weekly
$54.2 / week
Monthly
$100.3 / month
Quarterly
$272.8 / quarter
  • A good starting point for one app with 1–3 archives per day.
  • Start with sequential builds, then set concurrency based on measured memory use.
  • Add an SSD if you need to retain more build artifacts.
Rent Bake M4

Parallel testing & inference experiments

OSXBuild Bake M4 Pro

Extra memory and storage for multiple targets, simulator testing, and larger local models. Organize builds, tests, and artifact management around the same dedicated machine.

Chip
M4 Pro
Memory
64GB
SSD
2TB
$60.3 / day

Daily price shown. See the pricing below for other rental terms.

Daily
$60.3 / day
Weekly
$162.9 / week
Monthly
$301.6 / month
Quarterly
$820.4 / quarter
  • For multi-job builds, parallel simulator testing, and remote development.
  • Run local inference experiments with MLX / llama.cpp.
  • Test model size, context length, and concurrency against your workload.
Rent Bake M4 Pro

Choose how long to run these jobs

Use daily billing for short trials, weekly for a release push, and monthly or quarterly for ongoing CI. Switching updates the main prices on both cards and the selected rental term without hiding prices for other terms.

Daily pricing is shown. Both plans list prices for all four rental terms.

Compare configurations and pricing side by side

These are base model prices and exclude add-ons. Run a full archive with your own project and record peak memory, available storage, and queue times. That’s a better way to choose than comparing chip names alone.

ComparisonOSXBuild Bake M4OSXBuild Bake M4 Pro
Chip / memory / SSDM4 / 16GB / 256GBM4 Pro / 64GB / 2TB
Daily$20.1 / day$60.3 / day
Weekly$54.2 / week$162.9 / week
Monthly$100.3 / month$301.6 / month
Quarterly$272.8 / quarter$820.4 / quarter
Resource allocationWhole dedicated physical machine, not a VMWhole dedicated physical machine, not a VM
AccessmacOS desktop and command linemacOS desktop and command line
Choose where to build

2 Models Across All 6 Locations

Both configurations are available at the same locations. Choose based on your team’s working hours and where code and artifacts need to travel. If your team uses remote desktop often, test the connection from your own network before deciding.

Model availability by location
ModelSingaporeTokyoSeoulHong KongUS EastUS West
OSXBuild Bake M4AvailableAvailableAvailableAvailableAvailableAvailable
OSXBuild Bake M4 ProAvailableAvailableAvailableAvailableAvailableAvailable

“Available” means the model and location are listed for rental; it does not mean a machine is reserved. Check the console for live availability. All locations operate year-round, 365 days a year.

Location affects connectivity, not how resources are dedicated

For automated builds, consider the paths for fetching code, downloading dependencies, and uploading artifacts. For interactive development, focus on desktop responsiveness. The number of locations is not the number of concurrent builds a machine can run; base concurrency on actual project and test resource use.

Choose for your workload, not assumptions

Find the bottleneck before adding resources

These are starting points, not performance guarantees. Archive times depend on project size, dependencies, Xcode version, and cache state. Model compatibility also depends on quantization, context length, and runtime framework.

WorkloadTypical tasksSuggested starting pointWhat to check first
Light buildsOne app, 1–3 archives per day, sequential signing and export.M4 / 16GB / 256GBArchive times on the first run and with a warm cache, dependency size, and artifact retention needs.
Everyday developmentMultiple targets, parallel simulator testing, and coding over remote desktop.M4 / 16GB / 256GB + 1TB SSD, or M4 Pro / 64GB / 2TBTrack memory pressure and storage use separately. An SSD adds storage, not memory.
Large model inferenceRun medium to large models locally with MLX / llama.cpp.M4 Pro / 64GB / 2TBModel weights, runtime memory, context length, and acceptable inference speed.

Plan storage first

Estimate space separately for source code, dependencies, simulator runtimes, DerivedData, and archive artifacts. Move long-term IPA and log storage to your own artifact storage rather than treating the build machine as your only backup.

Measure before running jobs in parallel

Run one complete job first, then increase test or build concurrency. If memory pressure stays high, reconsider the model tier. If artifacts are simply accumulating, review retention policies and consider an SSD upgrade.

Make room for caches and artifacts

Optional add-ons, priced separately

Add an SSD or Thunderbolt 5 linking to your base configuration. Prices below are shown by rental term; linking is billed per machine. Confirm the final configuration and total when placing your order.

+1TB SSD

Add storage headroom for DerivedData, dependency caches, and simulator images. This add-on does not change the base model’s memory.

Daily
$2.1 / day
Weekly
$5.6 / week
Monthly
$10.4 / month
Quarterly
$28.3 / quarter

For workloads with a clear need for more storage, but not more memory.

+2TB SSD

For projects that need to keep more archives, test assets, and toolchain versions. Set up artifact cleanup and external backups before expanding storage.

Daily
$4.2 / day
Weekly
$11.2 / week
Monthly
$20.8 / month
Quarterly
$56.6 / quarter

Separate caches, regenerable data, and deliverables that must be retained.

Thunderbolt 5 linking

Priced per machine. For multi-machine workloads, check data transfer methods, software support, and deployment needs before enabling linking.

Daily · per machine
$1.8 / day
Weekly · per machine
$4.8 / week
Monthly · per machine
$8.8 / month
Quarterly · per machine
$23.9 / quarter

Linking does not combine memory automatically or distribute a single archive across multiple machines.

An SSD add-on is not a backup. Keep copies of source code, signing assets, and delivered artifacts somewhere you control, and export them before your rental term ends.

Review your handoff checklist before ordering

6 checks to avoid rework after setup

Confirm your workload, location, and delivery requirements. Have your SSH public key ready. Never include private keys, certificate passwords, or other sensitive information in support emails or ticket descriptions.

  • Workload defined

    Identify at least one primary goal: archiving, remote development, or inference experiments. Note project size and expected concurrency instead of relying on vague requests for “more performance.”

  • Location selected

    Prioritize your team’s connectivity and working hours. Interactive development and unattended builds have different requirements, so don’t choose based on distance alone.

  • Rental term and end date understood

    Confirm whether you chose daily, weekly, monthly, or quarterly billing, and review the order total before paying. After delivery, note the end date in the console and plan to renew or migrate.

  • SSD add-on evaluated

    Account for dependencies, simulators, and archives. If you choose an upgrade, confirm its capacity and price for your rental term. Don’t mistake add-on storage for the base configuration.

  • Payment method ready

    We accept USDT-TRC20 and Visa / Mastercard / Amex (via Stripe) only. All charges are in USD. Available payment gateways are confirmed by the backend at checkout.

  • SSH public key ready

    Keep your private key private; submit only your public key. After delivery, connect using the host, port, and username shown in the console. Don’t reuse connection details from another order.

Check payments and renewals in one place

Pay in USD and track orders in the console

Before paying, confirm the model, rental term, location, and add-ons. Use the price list to compare options; check the checkout details for the final amount and available payment gateways.

Payment options

USDT-TRC20 or Visa / Mastercard / Amex (via Stripe). For USDT-TRC20, follow the network, amount, and payment instructions shown with your order. Stripe processes card payments.

A payment screenshot alone does not confirm that an order is complete. After paying, return to the console to check the order status and delivery details.

Invoices, renewals, and end dates

View billing and invoice records in the console, and check the renewal option and end date for the relevant subscription. For long-running workloads, include renewal checks in your pre-release review rather than waiting for a build to fail.

When you’re done, stop the runner from accepting new jobs, export artifacts and required configuration, then prepare for the end of the rental term. Don’t leave your only copy of signing assets on a machine near its end date.

View billing in the console
A few things to know about your rental term

Get clear on configuration, renewals, and data

Choosing a plan is about more than the first invoice. Plan your migration and exit path too, so your build machine is ready for the next commit.

Can I upgrade to another model during my rental term?

To request an upgrade, open a support ticket in the console with your current order, target model, location, and preferred migration time. Replacing a physical machine isn’t the same as upgrading it in place: plan to migrate your toolchain, signing assets, and runner configuration.

The available options and costs depend on the order review. Back up your data before migrating, then verify everything on the new machine before switching the build queue. Make sure the old machine no longer accepts jobs.

Is the machine released automatically when my rental ends?

Machines that aren’t renewed by the end of the rental term enter the reclamation process. Don’t assume you can keep using a machine after its end date. Check the end date, renewal status, and order progress in the console.

Renew ahead of time if you need to keep building. Otherwise, export your code, artifacts, and required configuration before the end date. After reclamation, the disk is securely erased and the system reinstalled before the machine is assigned to another customer.

How do I keep my code and archived artifacts?

Code and artifacts on the machine belong to you. Commit source code to a repository you control, and sync any IPA files, logs, and other deliverables you need to keep to storage you control.

Before your rental ends, export any signing assets you need and check that your copies work. Regenerable data such as DerivedData can be recreated as needed; it is not a substitute for backing up source code or artifacts.

Which rental term should I choose when setting up CI?

First complete an end-to-end workflow: fetch code → install dependencies → archive → sign and export. Then choose a longer rental term. For a short trial, consider daily billing; for ongoing use, compare weekly, monthly, and quarterly prices against your expected usage.

Changing the billing term on this page only updates the displayed prices and the preselected term for new orders. It does not change the term of an existing subscription. Manage existing orders in the console.

Can an SSD upgrade fix slow builds or low memory?

An SSD upgrade adds storage capacity, not memory. First identify whether the bottleneck is repeated dependency downloads, low disk space, memory pressure, or a test queue, then choose resources accordingly.

If your workload needs more memory, consider M4 Pro / 64GB / 2TB. If caches and archives are simply accumulating, review your retention policy first, then decide whether to add an SSD.

Run your next archive on a machine of your own

Choose a model, rental term, and location, then review add-ons. Still deciding? Tell us about your build volume and project.

Rent a Cloud Mac Check rental and build support