Dedicated physical Mac · Not a VM

Choose your M4 server

Choose Flow M4 for lightweight, single-threaded builds; choose Flow M4 Pro for parallel builds and MLX experiments. Pick from two models and five locations, with daily, weekly, monthly or quarterly billing. Start with the workload, then choose the rental period that fits.

From $20.2/day. Check the portal for live availability. All prices are in USD.

Showing monthly pricing. Amounts are base rental prices per machine, excluding add-ons.

Flow M4

Move build workloads off your local machine.

M4
chip
16GB
unified memory
256GB
SSD

$100.8

/month · per machine · USD

For sequential iOS builds, React Native or Flutter projects, and a single self-hosted runner. Test your compile times and cache usage with your own project before committing to a longer billing period.

  • A dedicated physical Mac, not a VM. The entire machine is yours.
  • Full macOS desktop and command-line access, with VNC and SSH support.
  • Choose from five available locations to suit your team and working hours.
Rent Flow M4

Flow M4 Pro

More room for parallel builds and memory-intensive experiments.

M4 Pro
chip
64GB
unified memory
2TB
SSD

$306.1

/month · per machine · USD

For parallel builds, XCUITest regression testing and MLX quantised model experiments. 64GB of unified memory gives you more room to work, but validate concurrency against your project dependencies and model requirements.

  • A dedicated physical Mac, not a VM. The CPU, GPU and unified memory are not shared with other tenants.
  • Includes a 2TB SSD, so you can manage projects, models and caches separately.
  • Storage upgrades and Thunderbolt 5 interconnect options are available. Confirm the setup when you place your order.
Rent Flow M4 Pro

Compare all four billing periods

Changing the period above only changes the prices shown; it does not place an order. The full price list is below to help you plan with your team. Confirm the model, location, billing period and add-ons at checkout.

Base rental price per dedicated physical machine, in US dollars
Model and base specsDailyWeeklyMonthlyQuarterly
Flow M4
M4 / 16GB / 256GB
$20.2$54.4$100.8$274.2
Flow M4 Pro
M4 Pro / 64GB / 2TB
$61.2$165.3$306.1$832.6

Still validating your workload? Start with a daily rental to check your toolchain, memory use and remote access. Run a full build or inference task instead of choosing a model based only on idle system usage.

Have a steady workload? Weekly billing can suit release cycles; compare monthly and quarterly billing for always-on runners or ongoing experiments. Monthly pricing is based on 30 days and is not simply the daily rate multiplied by 30.

Add-ons for specific bottlenecks

Add storage when you need more capacity; consider interconnect when your work can use multiple machines. Add-ons are priced separately from the base machine. Check capacity, quantity and billing period before ordering—a larger disk is not a memory upgrade.

+1TB SSD

Extra room for build caches, dependency downloads and medium-sized projects. Check peak usage across your working directories and caches before adding capacity.

Daily
$2.7
Weekly
$7.2
Monthly
$13.3
Quarterly
$36.2

+2TB SSD

For model files, multiple build branches and larger test artefacts. Extra capacity is not a backup; keep separate copies of important data.

Daily
$5.4
Weekly
$14.4
Monthly
$26.6
Quarterly
$72.4

Thunderbolt 5 interconnect

For workloads spread across multiple machines; charged per machine. It does not combine several machines into one pool of unified memory, and your app must support splitting work across them.

Daily / machine
$1.9
Weekly / machine
$5.1
Monthly / machine
$9.4
Quarterly / machine
$25.6
Add-on prices for four billing periods, in US dollars; interconnect is priced per machine
Add-onDailyWeeklyMonthlyQuarterly
+1TB SSD$2.7$7.2$13.3$36.2
+2TB SSD$5.4$14.4$26.6$72.4
Thunderbolt 5 interconnect / machine$1.9$5.1$9.4$25.6

At checkout, check which add-ons are selected and confirm the number of machines using the interconnect. Before expanding, removing or resizing storage on an existing instance, check the portal for available options and when changes take effect. Data migration and billing changes are separate steps.

Both models available in all five locations

Singapore, Japan (Tokyo), South Korea (Seoul), Hong Kong and the US East Coast. Both models are listed in every location. Choose one close to your main operators; for CI-only workloads, also consider how code, dependencies and build artefacts will travel.

Models listed by location; this does not indicate live availability
ModelSingaporeJapan (Tokyo)South Korea (Seoul)Hong KongUS East
Flow M4AvailableAvailableAvailableAvailableAvailable
Flow M4 ProAvailableAvailableAvailableAvailableAvailable

Choose a location that suits your team. The table shows listed locations; check the portal for live availability. Remote desktop performance also depends on your local network, resolution and encoding settings, so location alone is not a guarantee of end-to-end performance.

A bill is more than one total

Here’s an example for one monthly Flow M4 Pro in Tokyo, with a +2TB SSD and one per-machine Thunderbolt 5 interconnect. This is an illustration of how charges add up, not a placed order.

Configuration and cost breakdown

Example · USD
Billing periodMonthly · 30 days

Pay in advance for the selected period. Check the period and order total before paying; you can see the expiry date in the portal.

Flow M4 Pro · Japan (Tokyo)$306.1

Base rental price for one dedicated physical machine with M4 Pro / 64GB / 2TB. Add-ons below are charged separately.

+2TB SSD$26.6

Extra storage is billed separately. Before resizing an existing instance, check available options, when the change takes effect and whether data needs to be moved.

Thunderbolt 5 interconnect ×1$9.4

This example is for one machine. Check the quantity for each machine in a multi-machine setup; the per-machine price is not the price for the whole cluster.

Example total $342.1 USD

Calculation: $306.1 + $26.6 + $9.4 = $342.1.

Before expiry, check your renewal order and the next billing period in the portal. If you change the configuration or remove an add-on, the charges and effective date confirmed in the portal will apply. This example does not describe mid-term refunds or prorated credits.

Check the currency and payment method

All listed prices, orders and payments are in US dollars (USD). Only the two payment methods below are supported. The payment gateway shown at checkout depends on the options returned by the payment service.

USDT-TRC20

Before sending, check that you’re using the TRC20 network and verify the receiving address and amount shown on the payment page. Make sure you select the right network, even if the asset name is the same.

On-chain confirmation times affect when your order is confirmed. After paying, check the order status in the portal. If you need help verifying a payment, keep your order number and transaction hash and provide them in a support ticket.

Visa / Mastercard / Amex

Card payments are processed by Stripe. Complete the payment verification shown at checkout, then return to the portal to check your order and machine delivery status.

If your bank shows a charge but your order status has not updated, first check that it relates to the same order, then submit a support ticket. Don’t make repeated payments just because the page has not updated yet.

For billing questions, include your order number, model, billing period and payment status. Never send your full card number, security code, machine password or SSH private key by email or support ticket.

Rental periods, renewals and changes

Plan your rental period and data handling first. If a machine is running, confirm your change plan before stopping workloads or moving files.

What happens to the machine and data if I don’t renew?

Don’t rely on a machine after its rental period as ongoing storage. When the rental period ends, the machine is reclaimed and its entire drive is erased before it is prepared for another customer. If you need to keep using it, check your renewal order in the portal before expiry. If not, export your code, certificates, models, build artefacts and any logs you need in advance.

Check the order and instance status in the portal to confirm a renewal; a transfer record alone is not enough. Keep independent backups of important data during your rental period and don’t count on recovering it after the machine is reclaimed.

Can I move my machine to another location mid-rental?

Each location corresponds to a physical machine; moving does not simply change a location label on the same machine. If you need to change locations mid-rental, submit a support ticket through the portal to confirm the available options, costs and timing. Don’t assume your running system, network address or data will move automatically.

Before migrating, save your project, dependency notes and build configuration. Once the new machine is verified, update your CI labels or team connection details. This keeps the change within steps you can check.

Weekly or quarterly: which should I choose?

Weekly billing can suit a defined release, testing or short-term delivery cycle. Quarterly billing is a better fit for ongoing workloads when you’ve already validated the model, location and toolchain. If you haven’t run a complete project yet, start with a daily rental and compare longer periods afterward. Don’t choose concurrency based on price alone.

Consider workload duration, data retention and how often your team will use the machine. This page lists prices for all billing periods. Before switching to the next period, confirm the available period and order total in the portal.

Can I remove add-ons or resize midway through a rental?

Check the portal for the changes available for your instance. For removals, resizing, refunds or billing adjustments on an existing order, confirm the applicable terms through a support ticket. Add-on prices on this page are for planning; a change is not guaranteed to take effect immediately whenever you request it.

Move files off an expanded drive before removing it, and stop related workloads before changing a multi-machine interconnect setup. After the change is confirmed, check mount paths, CI cache locations and app settings so your workloads don’t keep referencing resources that are no longer available.

Not sure what your project needs? Where should I start?

Record peak memory use, disk usage and the number of tasks running at once during a complete run. For lightweight, sequential builds, start by testing Flow M4. For parallel builds or MLX experiments, consider Flow M4 Pro. Storage upgrades add capacity, not unified memory.

Still unsure? Email support@osxflow.com with your use case, concurrent workload and estimated rental period. If you already have an order, sign in to the portal and submit a support ticket so the team can check your specific machine.

Ready to get started?

Choose a model, location and billing period, then check your add-ons. Get this workload running first; you can adjust your setup at renewal based on actual usage.