Built for developers and engineering teams in South China & Southeast Asia, with always-online Apple Silicon physical nodes for iOS/macOS builds, self-hosted runners, MLX inference services, and remote graphics workflows.
All three configurations provide a dedicated physical machine, not a virtual machine. Standard combinations are generally available to order; actual availability is returned in real time by the console.
All prices are in USD. Use daily or weekly billing for short-term validation, or monthly or quarterly billing for ongoing build and inference workloads. Each plan card shows the full price for every term, so you can see the total cost instead of calculating it from an entry price.
M4 · 16GB · 256GB
$19.2/day
Ideal for lightweight Xcode builds, automation scripts, small MLX model validation, and short-term compatibility testing.
M4 · 24GB · 512GB
$39.4/day
Built for continuous CI, React Native builds, medium-sized model inference, dependency caching, and rotating multiple automation tasks.
M4 Pro · 64GB · 2TB
$60.8/day
Designed for high-memory MLX inference, quantization validation for larger models, parallel runners, long-running builds, and archiving large artifacts.
This figure helps teams estimate the access distance for interactive development, remote graphics, and API calls. Actual latency depends on the access location, carrier routing, local network quality, encrypted tunnels, and task concurrency. Before migration, run SSH and graphical interface tests from the networks your team uses most.
For build queues, runners, model inference, and scheduled automation, a stable task execution path is usually more important than the lowest latency in a single test. Keep source code, dependency caches, build outputs, and model files on the same physical node to reduce repeated cross-network transfers.
Access baseline for South China & Southeast Asia to support initial region selection; it does not replace your team’s own network testing.
Estimated times indicate how long users typically need to complete each action and are not fixed delivery commitments. Actual availability, payment gateway, and node information are returned in real time by the console.
Choose among 16GB, 24GB, and 64GB based on peak memory, working-directory size, and concurrent task volume. For model inference, leave headroom above peak memory; for builds, check cache and artifact capacity as well.
Choose daily or weekly billing for short-term validation, or monthly or quarterly billing for stable pipelines. If your repository, dependency cache, models, and build artifacts may exceed the base disk capacity, review storage expansions when placing the order.
After confirming the machine, term, expansions, and total USD amount, pay with USDT-TRC20 or Visa / Mastercard / Amex (via Stripe). The actual available gateway is determined by the backend response.
Once the console returns the machine details, save the order number and connection parameters, then verify access via SSH or the macOS graphical interface. After your first login, update the password or keys and document the access boundaries your team will use going forward.
These answers focus on node selection, pricing, and order regions, without repeating connection troubleshooting guidance.
They are best suited to teams whose core members, code repository entry points, or upstream API users are located in South China & Southeast Asia and want shorter day-to-day remote interaction paths. For unattended builds or batch inference, also compare the locations of your source, artifacts, and callers rather than looking only at single-request latency.
The three base configurations on this page start at $19.2/day, $39.4/day, and $60.8/day, with weekly, monthly, and quarterly billing priced as shown on each card. Storage expansions and Thunderbolt 5 parallel connections are added separately to the order; the final amount is confirmed on the console checkout page.
The region is part of the order configuration, so a cross-region change should not be treated as an instant switch for the existing machine. If your workload needs a different region, first export your code, build artifacts, models, and required configuration, then create an order in the target region. After validating the new machine, revoke keys and tokens on the old one.
Check real-time configuration availability in the console before completing payment in USD. If you need to compare memory, storage, and all four billing cycles item by item, open the full plans page first.