Dedicated physical node · Not a VM

Rent a Hong Kong Cloud Mac
Starting with Mac mini M4

For developers in South China & Southeast Asia, run Xcode builds, CI jobs, and remote desktop sessions on one dedicated Apple Silicon machine. Choose a rental period that fits your project—no need to buy local hardware upfront.

Shell M4 from $21.8/day. Check provisioning status, connection details, and expiry date in the console.

  • 3 dedicated configurations
  • 4 billing periods
  • Available 365 days a year
Choose for your workload, then your rental period

One physical machine, rented for the time you need

All three configurations give you exclusive use of the machine's CPU, GPU, and unified memory. Start with daily rental for a quick test, then compare weekly, monthly, or quarterly options for ongoing builds. Changing the billing period only updates the displayed price and order selection—not the machine specifications.

Showing daily prices. All prices are in USD.

Shell M4

Single-project development and short build tests

$21.8/day

Chip
M4
Unified memory
16GB
Local SSD
256GB

Ideal for getting an Xcode build pipeline running, fixing a release issue, or checking whether an existing script works on macOS. Start with one build at a time, then decide whether to move up based on actual memory use.

  • Full SSH and macOS GUI access
  • Dedicated physical machine; hardware is not shared with other tenants
  • You manage your data, toolchains, and build artifacts
Rent Shell M4

Shell M4 Plus

Continuous integration and larger project caches

$39.6/day

Chip
M4
Unified memory
24GB
Local SSD
512GB

Designed for keeping dependency caches, working across multiple repositories, and connecting the machine to a self-hosted runner. Measure a clean build and a cached build before setting concurrency, so simultaneous builds don't exhaust memory or disk space.

  • Keep dependencies, DerivedData, and artifacts in separate directories
  • Compare monthly and quarterly billing for persistent team workloads
  • You control build logs and cache cleanup
Rent Shell M4 Plus

Shell M4 Pro

Heavier builds and unified memory workloads

$61.6/day

Chip
M4 Pro
Unified memory
64GB
Local SSD
2TB

Suitable for larger project workspaces, demanding builds, and MLX or llama.cpp experiments. Whether a model fits also depends on quantization, context length, and runtime overhead—not just the model file size.

  • More local storage for toolchains and experiment data
  • Run Metal workloads directly on the physical machine
  • Test your actual workload before committing to a longer period
Rent Shell M4 Pro

Need extra storage or a multi-machine setup? Check the order configuration for +1TB SSD, +2TB SSD, and Thunderbolt 5 daisy-chain add-ons. Add-ons have separate pricing by billing period and are not included in the machine rental prices above.

We accept USDT-TRC20 and Visa / Mastercard / Amex (via Stripe). All charges are in USD. Available payment gateways and node status are shown in real time in the console.

Understand the network figures first

What to expect from Hong Kong at your location

Average latency guidance for South China & Southeast Asia is 20~40 ms. Use this range to compare locations; it is not a measurement of your connection or a guarantee of the latency you'll see on every session.

Round-trip latency guide

20~40ms

Users in South China & Southeast Asia can consider Hong Kong as a potential node. This figure helps indicate interaction distance; your results will also depend on your ISP, international routing, Wi-Fi, and any local proxy.

For remote desktop, check latency and jitter. For file transfers, also consider throughput and packet loss—one figure can't tell you both.

Test it with your own connection

  1. Connect over SSH from your usual work network and check that command responses are consistent; don't rely on a single speed test.
  2. Upload a representative project dependency or test file, time the transfer, and see whether syncing large files fits your workflow.
  3. If you need a GUI, start at a lower resolution and adjust image quality gradually. This helps distinguish a local bandwidth bottleneck from a machine performance issue.

Nodes are available 365 days a year. Test network quality on your own connection, and track build time separately from time spent waiting on the network.

If your workflow is mostly committing code and waiting for CI, focus on repository checkout, dependency caching, and artifact uploads. If you spend hours at the desktop each day, test the interactive experience first. Don't make the final choice for your team based on geography alone.

Four steps to your first session

From placing an order to running your first command

The times below are estimates for the setup tasks, not provisioning guarantees. Payment confirmation, node delivery, and environment setup are separate steps; check the console for the latest status.

  1. Choose a configuration and period

    Estimated time: 2–3 min

    First, identify whether you need the machine for builds, CI, or inference. Then choose a configuration, billing period, and any add-ons. Before submitting, check the machine specifications, node, and total. Avoid paying for more concurrency than you've tested.

  2. Review your order and pay

    Estimated time: 1–3 min

    Choose a payment gateway available on the page. For USDT-TRC20, verify the network, payment details, and amount. For card payments, complete the verification steps through Stripe. Confirmation time depends on the payment result.

  3. Wait for delivery and get your credentials

    Details check: about 1 min

    Once delivery is complete, check the IP address, connection details, login credentials, and expiry date in the console. The order status shows when delivery is complete; a payment confirmation page does not mean the machine is ready to connect.

  4. Connect and check your toolchain

    Basic checks: 5–10 min

    On your first connection, verify the host fingerprint, update the login password, and configure an SSH public key. Then check the Xcode path and available disk space. Run a small project to confirm the build works; installing large toolchains and downloading projects take additional time.

Have these three things ready before you order

  • Know which toolchains you'll run, how large your projects are, and how much concurrency you expect.
  • Choose a billing period that covers your project, and plan to copy your artifacts before the rental expires.
  • Have a supported payment method ready so payment verification doesn't eat into your machine time.
Check these three details before you decide

Ordering a Hong Kong node

Choose your node before planning your workflow. These three points affect your first order and any future migration.

How can I confirm that my order will be delivered on a Hong Kong node?

This page covers Hong Kong, but you still need to check the selected node in the configurator. Don't rely on the page you came from. Before paying, confirm the node, configuration, billing period, and total shown in the configurator. Once the machine is delivered, check its instance details. The console shows current availability when you place your order.

I'm in South China & Southeast Asia. Will I always get 20~40 ms latency?

No. Regional guidance isn't a measurement of your individual connection, and 20~40 ms is a starting point for comparing locations—not a latency guarantee. Test SSH response times, desktop interaction, and file transfers on your usual work network and at the times you'll use the machine. Switching ISPs or proxies from the same location can also change your experience.

If our team changes office locations, will our existing workspace move automatically?

Don't plan on an automatic migration. Before switching nodes, confirm the new order and environment, then sync your code, required data, certificates, and build configuration yourself. Recheck your runner and job tags. Back up the old workspace before it expires, and avoid running the same release task on both machines during the migration.

ssh → build → ship

Start with one machine. Test it on a real workload.

Get started with Shell M4 from $21.8/day. Choose a configuration and billing period, confirm the node in your order, then proceed to payment. Base your choice on a real project, not a guess.