Dedicated physical Mac miniStart by the day. Deploy nearby.

Every order gives you a real Mac mini in a data centre—not a VM—with the CPU, memory and storage all to yourself.
For Mac mini hosting and mac mini server workloads alike: CI builds, device testing and local inference.

Specs · GNU Core M4
M4 · 16GB / 256GB
Ideal forLight builds / everyday development
Locations4 to choose from · Available
Session log · TokyoExample workflow
Machine readyGNU Studio M4 · System initialised
Remote loginScreen sharing and SSH public key
First buildXcode Archive · Check the build and logs
Illustrative timings only; not a delivery or build-time commitment
4 locations · One price
SingaporeSGAvailable
TokyoJPAvailable
SeoulKRAvailable
Hong KongHKAvailable
Specs · GNU Studio M4
M4 · 24GB / 512GB
Ideal forLarge-model inference / multiple simulators
Add storage+1TB or +2TB SSD
From · GNU Core M4
$21.3/day
Week
$57.5
Month
$106.4
Quarter
$289.4

Start with your workload

Choose a workload to highlight a recommended option below. All configurations remain visible.

Dedicated physical machinesNot VMs4 Asia-Pacific locationsDaily / weekly / monthly / quarterlySettled in USDUSDT-TRC20 or Visa / Mastercard / Amex via Stripe

Use cases

Four teams, four ways to work

Here are common workflows in the order you’d use them. Decide what you need to deliver, then choose a configuration, location and rental term. You’ll need to provide your own tool licences and project dependencies.

Indie developer

Build overnight, reach TestFlight by morning

  1. Write code locally during the day and push the project and dependency versions to a remote repository.
  2. Log in to the Tokyo location in the evening, configure your signing assets and run Xcode Archive.
  3. Check the build, upload it to TestFlight, and save the build logs and exported files to team storage.

For a single project, start with the 16GB configuration. Have your certificates, provisioning profiles and distribution permissions ready; renting a machine doesn’t provide project licences.

GNU Core M4TokyoMonthly
CI team

Run signing and tests on a self-hosted runner

  1. Register a self-hosted runner in Singapore and limit its repository and job permissions.
  2. Use pull requests to trigger builds, code signing and unit tests, with consistent logs for every run.
  3. Rent by the week to cover a sprint, and archive caches separately from release builds.

For parallel jobs, check memory pressure before setting concurrency. A dedicated machine avoids contention with other tenants, but won’t replace project-level cache optimisation.

GNU Studio M4SingaporeWeekly
AI research

Compare local inference runs, one day at a time

  1. Set up an MLX environment on the 24GB configuration and pin your runtime and model versions.
  2. Load a quantised model and record context length, memory use and inference results.
  3. Export your scripts and experiment notes, then back them up and return the machine before the rental ends.

Whether a model fits in memory depends on quantisation, context length and runtime overhead. Extra SSD adds file storage, not unified memory.

GNU Studio M4SeoulDaily
Audio and video

Batch transcode and transfer finished files over SFTP

  1. Upload your media to the Hong Kong location. Add an SSD if you have a large media library.
  2. Set up your own Final Cut / Compressor environment and run batch transcodes.
  3. Check the finished files, then transfer them to local or team storage over SFTP.

Test your encoding settings and transfer time with a representative clip first. Keep source media, temporary files and output in separate directories to avoid filling the disk mid-job.

GNU Studio M4Hong Kong+1TB SSD
Pricing

Two configurations, with prices upfront

Both options are full Mac mini M4 machines, with the same price at all four locations. Prices are in USD. Start with a day to test your workload, then choose a week, month or quarter if you need longer.

GNU Core M4

M4 · 16GB RAM · 256GB SSD

$21.3per day · from

For single-project builds, unit tests and everyday remote development. On your first run, check build time, memory pressure and project cache size.

Day
$21.3
Week
$57.5
Month
$106.4
Quarter
$289.4
Singapore / Tokyo / Seoul / Hong Kong · AvailableChoose this

GNU Studio M4

M4 · 24GB RAM · 512GB SSD

$40.6per day

For workflows with larger model files and multiple simulators running in parallel. Choose more memory and local storage when needed; extra storage isn’t a compute upgrade.

Day
$40.6
Week
$109.7
Month
$203.2
Quarter
$552.7
Singapore / Tokyo / Seoul / Hong Kong · AvailableChoose this

Add-ons+1TB SSD $2.7/day and up+2TB SSD $5.4/day and upThunderbolt 5 linking $1.3/day and up (per machine)

Prices shown are catalogue prices. Availability, add-on combinations and the checkout total are confirmed in real time in the dashboard. Thunderbolt 5 linking does not automatically combine the memory of multiple machines.

Choose a setup

Pick a configuration for your workload

Start with the workload, then choose a configuration. Memory affects how many simulators you can run at once and how large a model you can load. SSD capacity determines how much room you have for projects, caches and media.

Workload, typical load, recommended configuration and extra storage
Workload Typical load Recommended Extra storage
Light builds Single-project Xcode builds and unit tests GNU Core M4 16GB / 256GB Start with the built-in SSD; check actual project and cache usage
Everyday development Solo development and debugging, 1 to 2 simulators GNU Core M4 16GB / 256GB Add +1TB SSD if you have large caches
Large-model inference Quantised models with MLX or llama.cpp GNU Studio M4 24GB / 512GB Add +1TB or +2TB SSD for multiple model files
Multiple simulators Multi-simulator UI tests, Unity or Flutter projects GNU Studio M4 24GB / 512GB Add +1TB or +2TB SSD for larger media files

Not sure? Rent a machine for a day and test your workload. Record peak memory use, cache growth and build size before choosing a rental term or extra storage. Test with a real project, not an empty one.

See all plans and recommendations
Locations

4 Asia-Pacific locations, close to your team

Choose a data centre based on where the people connecting to the machine are—not just where your project’s users are. All four locations have the same pricing, and both configurations are available. Check live availability in the dashboard.

SG

Singapore

A good fit for development and collaboration teams in Southeast Asia. If you run a CI runner here, check the route to your code repository and build storage.

  • Core · 16GB / 256GBAvailable
  • Studio · 24GB / 512GBAvailable
Location details
JP

Tokyo, Japan

For developers and localisation testing teams in Japan. A location doesn’t change distribution platform review requirements or guarantee an outcome.

  • Core · 16GB / 256GBAvailable
  • Studio · 24GB / 512GBAvailable
Location details
KR

Seoul, South Korea

A good fit for game and app teams targeting South Korea. Test your connection before remote debugging, and check transfer times when running batch tests.

  • Core · 16GB / 256GBAvailable
  • Studio · 24GB / 512GBAvailable
Location details
HK

Hong Kong

For teams across the region working together or running centralised builds. If several people share a machine, agree on a job queue and separate working directories.

  • Core · 16GB / 256GBAvailable
  • Studio · 24GB / 512GBAvailable
Location details

All locations operate year-round, 365 days a year. Screen sharing is more sensitive to round-trip latency, while SSH and unattended builds rely more on connection stability. Test your local network before scheduling production work.

“I moved my overnight builds to the Tokyo machine. By the time I get to work, the build is already in TestFlight.”
Indie iOS developer
“With a dedicated machine, the simulators aren’t slowed down by other tenants. Our CI queue is noticeably shorter.”
Mobile CI lead
“I rent the 24GB setup for a day to compare inference runs, then return it the same day. The billing is easy to track.”
AI research engineer
Get started

Up and running in three steps

Place your order, pay and retrieve your credentials in the dashboard. Once your machine is ready, check the connection and disk before moving over a production project. Keep another copy of your code and signing assets somewhere other than the rented machine.

01

Choose a configuration and location

Pick GNU Core M4 or GNU Studio M4, then choose Singapore, Tokyo, Seoul or Hong Kong. Check the billing period, extra storage and number of machines, and review every amount in your order.

What you getUnpaid order with your configuration, location, billing period, add-ons and USD total.
02

Pay in USD

Pay with USDT-TRC20 or by card using Visa / Mastercard / Amex via Stripe. Follow the payment page instructions. Available payment gateways are confirmed by the backend.

What you getPayment confirmation and order number; verify USDT payments against the on-chain record.
03

Get your credentials and log in

Once setup is complete, retrieve the host address and initial credentials from the dashboard. Connect via screen sharing or SSH, change the initial password and add your own public key, then run a first verification task.

What you getRemote access details and a macOS environment ready for development and builds.
Rent a Mac mini Connection guide
FAQ

Frequently asked questions

Here are five questions people often ask before ordering. Remote access doesn’t bring your local peripherals into the data centre. For iOS device testing, check the guide for network requirements and device connection limits first.

Browse all guides
Is this a virtual machine?

No. Every order is for a dedicated physical Mac mini. Its CPU, memory and storage aren’t shared with other tenants, and the machine isn’t split into virtual machines. You get the full macOS graphical interface and command line, so you can set up your own build toolchain.

Dedicated means exclusive use of the hardware; it isn’t a performance guarantee for every task. Dependency downloads, build caches, model quantisation and network routes can all affect completion time.

Can I upgrade my configuration mid-rental?

If you need more memory or built-in storage, sign in to the dashboard to request a change. Confirm the target configuration, cost and migration plan. Changing configurations involves another physical machine; it doesn’t add memory to your current machine on the fly.

Back up your project, environment settings and build files before migrating, then verify your jobs. If you only need more file space, consider a +1TB or +2TB SSD instead of treating a storage issue as a memory issue.

How do I renew my rental?

Choose from four billing periods: day, week, month or quarter. Check the expiry time, renewal options and amount due in the dashboard, and confirm the payment status. Starting a payment doesn’t necessarily mean your renewal is complete.

Renewal options, reminders and order status are shown in the dashboard. If you no longer need the machine, export your data, stop your jobs and remove project credentials before it expires. Don’t use an expired machine as backup storage.

Which payment methods do you accept?

All payments are settled in US dollars (USD). We accept USDT-TRC20 and Visa / Mastercard / Amex processed by Stripe. Available payment gateways are confirmed by the backend.

Card payments are processed by Stripe; we don’t store full card numbers. For USDT, check the network, recipient details and amount, and keep the transaction record. For payment queries, submit a support ticket through the dashboard with your order number.

What happens to my data when I return the machine?

At the end of the rental, the machine’s disk is erased and the system is reinstalled. Any added +1TB or +2TB SSD is erased as well. The machine is reset before it’s handed to the next user, so the previous user’s data isn’t retained.

Erased data can’t be recovered. Back up your code, models, logs and build files before returning the machine, and revoke runner registrations and any project keys you no longer need. See the data security guide for details.

Data isolation and deletion details
Engineering notes

Turn each session into a reusable record

Follow real-world workflows, from builds to operations. Pin your dependency versions before you start, then save the logs when you’re done so you can reproduce the work next time.

DevOps & CI/CD

Test iOS geographic distances and geofence boundaries on a cloud Mac

Build deterministic distance checks with fixed coordinates, then separate antimeridian handling and boundary decisions from real-device location behavior.

 DevOps & CI/CD

Test iOS WebSocket heartbeats and isolate connection generations

Separate heartbeat policy from transport code, then use explicit time and connection generations to test deadline boundaries and reject stale callbacks.

 AIDevelopment

Testing iOS progress weights, cancellation and completion

Build deterministic tests for weighted progress and cancellation, then keep counter completion separate from successful delivery.

Choose a configuration and location. Move your next build to the cloud.

Get started Read the guides