Start with the work. Choose the right Mac.
A cloud Mac gives you an entire physical Mac mini to yourself—not a virtual machine carved out of shared hardware. Use it for iOS builds, a macOS desktop or a focused stretch of compute for a short-term experiment. Start with the task, then choose a configuration, location and rental period: daily, weekly, monthly or quarterly.
Two M4 configurations · from $21.4/day · ready in about 4 minutes. Availability is confirmed in real time in the console.
Match the workload, then choose the term
From a single build to a reliable build queue.
Let the task shape how you use your Mac
All five use cases start with the same dedicated Mac, but each calls for different files, connection methods and rental periods.
Move iOS builds out of the shared queue
Register your cloud Mac as a self-hosted runner: a machine you manage and dedicate to CI jobs. Run PR checks and nightly builds on your Mac, without competing with other tenants for the same hardware.
Three things to do first
- Set up the build environment with the Xcode version your repository requires, then match the runner labels to your project jobs.
- Keep dependency caches and DerivedData in a consistent location to avoid downloading them again for every job.
- Limit access to signing materials to release jobs that need them. Keep failed-job logs in your team's CI system for troubleshooting.
Configuration For a single app, Fastlane and unit tests, start with the XV M4 Runner. For a short release push, try a daily or weekly rental. If you need to run more jobs in parallel, compare adding another machine with moving to Studio.
Archive and sign in a clean macOS environment
As a release approaches, the last thing you want is a toolchain that differs from the one you used last time. Run the archive, signing checks and App Store Connect upload on the same cloud Mac to make build records, certificate locations and artifact paths easier to verify.
Check each step in order
- Pin the required Xcode version and dependencies, then create a reproducible archive.
- Check that the signing certificate and provisioning profile match the target app. Never commit sensitive materials to your repository.
- Review the exported artifacts before uploading. Before your rental ends, retrieve any logs, archives and files you need to keep.
Configuration For packaging a single app, start with the XV M4 Runner. For multiple targets and Simulators running at once, choose the XV M4 Studio and reserve it for the full release cycle.
Build iOS apps from different frameworks in one place
React Native, Flutter and Unity development can happen on different operating systems, but final iOS builds still need the macOS toolchain. Bring each project's iOS builds together on a cloud Mac so your team can check artifacts against the same Xcode environment.
Get your inputs in order first
- Note the required framework, plugin and Xcode versions. Get a minimal build working before adding it to the full pipeline.
- Keep source caches, dependency caches and final archives separate. Clearing caches won't solve a disk capacity problem.
- Allow for enough memory and SSD space if you build several projects at once or debug them in parallel over remote desktop.
Configuration For sequential builds, check whether the Runner's 16GB of memory and 256GB SSD are enough. For multiple projects in parallel, consider the Studio's 24GB of memory and 512GB SSD.
Test local network performance from Asia-Pacific locations
Choose from four locations: Singapore, Tokyo, Seoul and Hong Kong. Select a location in your target users' region, run tests on that Mac, then compare the results with your local tests.
Include the location in your test notes
- Record the location, test device and network conditions. One connection test doesn't represent an entire region.
- Remote desktop responsiveness also depends on your distance from the Mac. Start interactive testing from a location near the person operating it.
- When comparing locations, keep the project version and test steps consistent. Both Mac tiers are available in all four locations; availability is confirmed in real time in the console.
Configuration For a single-location check, start with a daily Runner rental. For ongoing regression tests or multiple Simulators in parallel, consider Studio and a longer rental period.
Give Apple Silicon inference experiments their own Mac
Try MLX or llama.cpp with Apple Silicon unified memory without buying hardware just to run a short-term test. Whether a quantized model of around 7B–13B parameters fits your setup depends on the quantization method, context length and other workloads running at the same time.
Start small, then scale up
- Download one model and a small set of test prompts. Check actual memory use, free disk space and response times.
- Keep model weights, experiment outputs and environment records separate. Add a +1TB SSD if you need to store more weights.
- Export your results and any files you want to keep before the rental ends. The machine's disk is fully reset when the rental expires.
Configuration Start by testing with the XV M4 Studio's 24GB of memory and 512GB SSD. A daily rental is a good way to check feasibility; choose extra storage based on the size of your model weights.
Match your workload to a Mac and rental period
Use this table as a starting point, then measure your project's memory use, disk space and parallel workload. The catalogue has two M4 tiers; add-on storage and Thunderbolt 5 connections are optional.
| Workload | Suggested configuration | Start with | Why this fits |
|---|---|---|---|
| Light builds One app, Fastlane, unit tests |
XV M4 Runner M4 / 16GB / 256GB |
Daily or weekly | Get one complete pipeline running, then decide whether to extend the rental based on your actual queue length. |
| Day-to-day development Multiple targets, Simulators, remote desktop |
XV M4 Studio M4 / 24GB / 512GB |
Monthly | Run parallel debugging, build caches and desktop sessions on the same machine. |
| Quantized model inference Around 7B–13B; measure actual usage first |
XV M4 Studio M4 / 24GB / 512GB, optional +1TB SSD |
Try daily first | Use measured results to choose a model and context size. Extra storage holds model weights; it doesn't add unified memory. |
| Team build jobs Multiple projects or parallel pipelines |
Multiple XV M4 Studio machines Each: M4 / 24GB / 512GB, optional Thunderbolt 5 connection |
Match the project timeline | Assign runner labels and jobs first, then check whether the machines actually need a high-speed connection. |
Starting prices: Runner $21.4/day; Studio $40.7/day; +1TB SSD $2.4/day; Thunderbolt 5 connection $1/day per machine. See all rental plans for weekly, monthly, quarterly and other add-on prices.
Get started in four steps—no long-term commitment required
Not sure how much parallel capacity you need? Test it with a real build or experiment first. Before ordering, prepare any data you need to keep, your SSH public key and your project dependencies.
- 01
Choose a use case
Decide whether you mainly need CI, release builds, cross-platform packaging, regional testing or inference experiments, and estimate how long you'll need the machine.
- 02
Choose a tier
Pick Runner or Studio based on memory, disk space and parallel jobs. For regional network tests, choose from four Asia-Pacific locations.
- 03
Configure and order
Choose a daily, weekly, monthly or quarterly rental and any add-ons. Payment is accepted only in USDT-TRC20 or by Visa / Mastercard / Amex processed through Stripe; all prices are in US dollars.
- 04
Connect and test
Once you receive your access credentials, test SSH and remote desktop, then run a representative task. Record versions, run times and disk usage.
Rent for a day and get your first task running.
Start with the XV M4 Runner from $21.4/day. If you already know you need more parallel capacity or memory, choose the XV M4 Studio. Both give you an entire machine to yourself.