Choose a plan to match your workload
Rent a dedicated physical Mac mini and keep using your existing Xcode, fastlane and CI pipelines. It’s not a virtual machine, and its CPU, unified memory and SSD aren’t shared with other users. Starting at $20.5/day.
Pay upfront by the day, week, month or quarter. Prices below are for the full selected term, not a daily-rate calculation.
Switch terms to update all three plan cards; see the full price list below.
Showing the full daily-term price in USD.
Move single-app builds, uploads and routine archives off your local machine.
$20.5 / day
More headroom for everyday multi-target development, simulators and cross-platform builds.
$39 / day
More memory for large archives, Unity exports and MLX inference experiments.
$61 / day
Before choosing a longer term, consider how long you’ll need the machine. These are list prices for each model. Monthly prices aren’t broken down into daily rates, and add-ons aren’t included in the base rental price.
| Model | Day | Week | Month | Quarter |
|---|---|---|---|---|
| Host M4 Lite | $20.5 | $55.4 | $102.5 | $278.8 |
| Host M4 Plus | $39 | $105.4 | $195.1 | $530.7 |
| Host M4 Pro | $61 | $164.6 | $304.8 | $829.1 |
All prices are in US dollars (USD). Pay with USDT-TRC20 or Visa, Mastercard or Amex via Stripe. Available payment gateways are confirmed by the backend at checkout.
Every spec, clearly listed
Each plan gives you an entire dedicated physical machine. Check the listed hardware specifications first, then confirm your project’s required OS and toolchain versions.
MODEL / m4-16-256
MODEL / m4-24-512
MODEL / m4pro-64-2tb
Don’t choose by chip name alone.Build times for the same project also depend on compiler settings, dependency caches, signing workflows and concurrency. We don’t present unmeasured timings as performance guarantees. If you require a specific Xcode or SDK version, or have network requirements, confirm them through a support ticket in the console before planning your migration.
Choose for your bottleneck, not the name
Run a real project first, then check peak memory usage, cache size and job queues. These recommendations are a starting point, not a fixed concurrency guarantee.
| Workload | Recommended plan | Why it fits |
|---|---|---|
| Lightweight single-app builds and fastlane uploads | Host M4 Lite | Start with one job to validate dependencies and signing. 16GB of unified memory is a practical starting point for a lightweight pipeline; keep an eye on caches and build artifacts accumulating on the 256GB SSD. |
| Everyday multi-target development, parallel simulators, or the iOS stage of a Flutter pipeline | Host M4 Plus | 24GB of unified memory and a 512GB SSD provide more room for multiple targets, simulator runtimes and dependency caches. Keep the Android stage in your existing environment. |
| Concurrent archives for large projects, Unity exports, or large-model inference | Host M4 Pro | 64GB of unified memory and a 2TB SSD support heavier workloads. Measure actual memory use before deciding on archive concurrency or model size. |
Keep the project commit, Xcode version and build configuration fixed. Record first-build and cache-hit results separately; don’t count dependency download time as compile performance. If simulators and archives run together, monitor memory pressure. If storage is simply filling up, start with cache cleanup and storage planning instead of upgrading to a higher-tier chip.
Add-ons, priced separately
These are add-on prices, not a separate plan. The base SSD remains as specified in your chosen plan; confirm upgrade charges when configuring your order.
Add storage for dependency caches, multiple toolchain versions and build artifacts. First decide which directories need to be retained, then plan your migration and mounts.
Adds SSD storage, not unified memory.
For larger project assets, game exports or model files. Keep long-term artifacts separate from caches that can be rebuilt.
Back up expansion storage separately; it isn’t a substitute for archiving your build artifacts.
Priced per machine. For multi-machine setups, confirm the topology, ports on each model and how jobs will be divided. Linking machines does not combine their memory into a shared pool.
Available combinations and delivered configurations are confirmed in the console.
Storage upgrades address capacity; they won’t automatically speed up every build step. For linked machines, specify the number of machines, data flows and how jobs will be split before ordering. Don’t pay for separate orders and assume the machines will automatically work as a group.
3 plans, 5 locations
Singapore, Japan (Tokyo), South Korea (Seoul), Hong Kong and the US West. All 3 plans are offered at these 5 locations. Check the console for real-time availability.
| Model | Singapore | Japan (Tokyo) | South Korea (Seoul) | Hong Kong | US West |
|---|---|---|---|---|---|
| Host M4 Lite | Available | Available | Available | Available | Available |
| Host M4 Plus | Available | Available | Available | Available | Available |
| Host M4 Pro | Available | Available | Available | Available | Available |
For frequent GUI use, consider the network route between you and the machine. For CI-only workloads, also consider the locations of your code repository, dependency sources and artifact destinations. A nearby time zone doesn’t guarantee lower latency. We don’t publish untested latency figures for any city. Test connectivity, dependency downloads and artifact uploads before settling on a location for your team.
Runs year-round, 365 days a year
99.9%monthly uptime commitment
For a 30-day month, this calculation corresponds to about 43 minutes of downtime. This is the basis for calculating the service commitment, not historical monitoring data or permission to intentionally interrupt service.
For downtime beyond the allowance, service is extended by the equivalent number of days based on the duration of the impact. Include a clear incident timeline so we can verify the order and affected location.
Include your order number, location, start and recovery times, and time zone.
Include connection errors, job logs and troubleshooting results. Redact passwords, private keys and signing materials.
Submit your claim through the console within 7 days after the incident ends. Support will verify the affected period.
Plan payment and expiry in advance
Check the model, location, billing term and add-ons before paying. Payment and instance delivery are separate statuses; verify both in the console.
Orders are priced in USD. Before paying, check the network, payment details and amount due shown by the gateway. Don’t use payment details from another order.
To ask about a payment, include your order number, transaction hash, amount and payment time in a support ticket. Never share your wallet’s private key or unrelated credentials.
Follow the payment gateway details for your orderCard payments are processed by Stripe and settled in USD. Available payment gateways are confirmed by the backend. After confirming your order, follow the on-page instructions to pay.
If you have questions about payment status, check your statement and order status first, then submit your order number and any necessary payment evidence. Never send your full card number or security code in a support ticket.
Only supported card brands are listedBefore expiry, check the renewal amount and order dates in the console. Service that isn’t renewed ends according to the order terms. Data retention and release follow the order and service terms; back up your data before release.
Don’t rely on unconfirmed retention periods as a backup plan. Export build artifacts, configuration and any essential logs before ending your service. Machines are securely erased after rental ends.
Confirm expiry arrangements before payingCheck these 3 things before ordering
For changes to an existing order, confirm what’s possible and any fees before migrating data. Don’t assume unconfirmed changes take effect automatically.
Submit your current order number, target plan and preferred migration time in a support ticket through the console. A plan change involves different hardware for a dedicated physical machine; it doesn’t work like adding memory to a virtual machine while it’s running. Support will confirm whether a new order is needed, how the remaining term is handled and the migration sequence.
Back up project configuration, build artifacts and signing materials before migrating. Run an archive in the new environment before ending use of the old one.
Don’t remove an expansion drive that’s still used by your project, caches or model directories. First, submit a support ticket through the console with your order number, add-on and preferred end date, and confirm that your data has been moved. Whether the add-on can be changed separately and how any remaining fees are handled depend on the order terms and review. We don’t promise automatic prorated refunds.
When submitting your request, specify the mount point and what data needs to be retained. Don’t include private keys, passwords or certificate contents.
Each of the 4 billing terms has its own full-term list price; longer terms aren’t calculated by multiplying the daily rate. For example, Host M4 Lite costs $20.5/day, $55.4/week, $102.5/month or $278.8/quarter. The monthly price isn’t a “pay per day” amount, and the quarterly price isn’t the total of 3 monthly orders.
Actual start and end times, renewals and billing-term changes follow the order details. If you need to change terms, confirm the options in the console first; don’t calculate refunds yourself by the hour or by the remaining days.
List your project’s required Xcode and SDK versions, dependency managers and build commands, then check that they’re compatible with the target macOS version. Confirm specific version requirements in a support ticket before ordering to avoid finding out after payment that the toolchain combination won’t work as expected. The macOS GUI and command line are both available. See the connection and migration guide for environment setup details.
Give your next build a clear path
Start by testing with a real project. Confirm the model, location and billing term, then move routine builds from your local machine to a dedicated physical machine.