Dedicated physical node · Tokyo

Cloud Mac rental in Tokyo
Mac mini M4 for your build pipeline

For users in East Asia, run your projects, toolchains and build jobs on one dedicated Apple Silicon physical machine. macOS runs natively, not in a virtual machine. Run tasks over SSH and use the graphical interface for operations that need human input.

From $21.2/day · billed in USD
USDT-TRC20 or Visa / Mastercard / Amex (via Stripe)

Choose resources, then a term

Three configurations, four billing periods

Each plan comes with a fixed chip, memory and base SSD configuration. Shared compute resources are never presented as a dedicated machine. Prices below are in USD and charged once for the selected period. Add-ons cost extra; confirm the total when you place your order.

Showing daily pricing. You can still change the billing period at checkout.

FTP M4 Core

A good starting point for testing remote access, script compatibility and a single-project build workflow. If you already know how much memory you need, choose the right plan for that workload rather than using a longer term to make up for limited capacity.

$21.2 / day
Chip
M4
Unified memory
16GB
SSD
256GB
Daily
$21.2
Weekly
$57.1
Monthly
$105.8
Quarterly
$287.8
Rent FTP M4 Core

FTP M4 Plus

More memory and workspace for continuous integration and day-to-day development. Keep dependency caches and build logs, and connect your own self-hosted runner.

$39.7 / day
Chip
M4
Unified memory
24GB
SSD
512GB
Daily
$39.7
Weekly
$107.2
Monthly
$198.6
Quarterly
$540.2
Rent FTP M4 Plus

FTP M4 Pro

For builds that use more memory and AI inference experiments. Check model loading, context size and artifact storage before deciding how many tasks to run in parallel. Choose based on workload, not just the chip name.

$60.1 / day
Chip
M4 Pro
Unified memory
64GB
SSD
2TB
Daily
$60.1
Weekly
$162.2
Monthly
$300.3
Quarterly
$816.8
Rent FTP M4 Pro

Changing the billing period compares prices; it doesn’t place an order

Your selection on this page only changes the prices shown. In the configurator, check the plan, node, billing period and add-ons before confirming payment. Start with daily billing for a short test. For an ongoing project, compare the full weekly, monthly and quarterly prices rather than multiplying the daily rate to estimate the cost. You can renew before your term ends; always keep a separate backup of important files.

Latency is only one part of the connection

Tokyo node: check latency and your workflow

Estimated average latency for users in East Asia: 30~60ms. Use this as a guide when choosing a region. It is not a guarantee for every provider, session or app response time, and it is not a live speed test from this page.

Interactive work depends on round-trip latency

SSH input, remote desktop and frequent small-file operations are more sensitive to round-trip latency. Even with the same latency in milliseconds, packet loss and jitter can make the experience feel different. Test from your actual work network instead of judging by geographic distance alone.

Test during both work hours and the evening. Check whether connections are stable and the interface responds smoothly. If you use a corporate proxy or VPN, test over your final connection path.

Bulk jobs depend on the whole transfer path

Project uploads, dependency downloads and build artifact transfers depend more on actual throughput, remote service response times and file counts. The route from a Cloud Mac to your code repository is different from the route between your local machine and the Cloud Mac.

For the initial sync, transfer only what you need and exclude rebuildable directories such as DerivedData. Then resolve dependencies and run an archive build on the same project. Track network wait time separately from build time to find the bottleneck.

How to use the estimate when choosing a node

Test using the actual address provided in the control panel. ICMP ping can help you check round-trip latency and packet loss, but if ICMP is restricted, that alone does not mean SSH or graphical access is unavailable. This page does not include raw samples or claim the estimate comes from a week of evening peak-hour testing. Your network, protocol and actual workload are what matter.

Confirm the details before paying

From order to access in four steps

The times below are estimates for the manual steps involved. They are not guarantees of payment processing or machine setup times. Check payment, provisioning and credential status in the control panel. A full project migration is not included in these four steps.

  1. 01

    Check your plan

    In the configurator, confirm the plan, node, billing period and add-ons. Check memory needs, disk usage and project duration together rather than comparing only the entry-level daily price.

    Estimated hands-on time: 2–5 minutes
  2. 02

    Complete payment

    Check the amount due in USD, then choose USDT-TRC20 or Visa / Mastercard / Amex via Stripe. Available payment gateways depend on the response from the backend. Follow the instructions on the payment page.

    Estimated hands-on time: 2–5 minutes; processing time not included
  3. 03

    Check delivery

    Review your order and provisioning status in the control panel. Once the machine is ready, verify the instance details, connection address and access credentials. If the status is still pending, follow the order update and don’t pay again.

    Estimated review time: 1–2 minutes; provisioning time depends on order status
  4. 04

    Run a first test

    Connect over SSH or through the graphical interface. Change the access password, configure your keys and check the toolchain path. Run a minimal build and verify its logs, exit status and artifact location before migrating the full pipeline.

    Estimated basic test: 5–10 minutes; downloads and compilation not included
Choosing a node

Three things to check before choosing Tokyo

These answers cover deployment location and project fit, not connection troubleshooting.

Does 30~60ms reflect the experience of every user in East Asia?

No. This estimate helps you decide whether Tokyo is worth testing first; it is not a measurement for every user. Carrier routing, corporate network egress, proxies, wireless connections and time of day can all affect latency and jitter. For remote desktop, test how smoothly it responds. For a build pipeline, track dependency downloads, compilation and artifact transfers separately.

For app releases targeting East Asia, does the build machine have to be in Tokyo?

No. The app’s target market and the build machine’s location do not have to match. Choose a node based on your team’s remote access, code repository, dependency sources and artifact download path. App distribution requirements and build machine access are separate considerations; node location does not replace distribution workflows or project compliance checks.

Can I choose any of the three configurations in Tokyo? Does the term affect resource allocation?

The node catalogue includes FTP M4 Core, FTP M4 Plus and FTP M4 Pro. Check the control panel for live availability. Daily, weekly, monthly and quarterly terms change the billing period; they do not turn a dedicated physical machine into a shared virtual machine. The three configurations have fixed specifications. If you need more storage or parallel processing capacity, check the available add-ons and their prices in the configurator.

Choose your plan and get your first build running

Three fixed configurations and four billing periods. Check your project requirements, then confirm the node and amount due in the configurator. Once payment is complete, follow delivery in the control panel.