Configurations, Terms & Nodes

Choose a Cloud Mac for your workload—not idle hardware

VMKeep offers three Apple Silicon dedicated physical machine plans. Every order maps to a dedicated physical node—not a virtual machine—with daily, weekly, monthly, or quarterly terms for short-term validation, continuous builds, and high-memory MLX inference.

Rent a Cloud Mac now Review connection and migration

3 plans: Fixed configurations available
5: Available nodes or data centers
$19.2/day: Lowest starting price

VMKEEP / RENTAL DOSSIER Plan selection record

All plans use dedicated physical machines

Set resource limits first, then choose the term

Prioritize peak memory, storage usage, and task duration. Nodes shorten the connection path for your team without changing the three plan configurations.

Light validation: M4 / 16GB / 256GB
Continuous workflows: M4 / 24GB / 512GB
High-memory inference: M4 Pro / 64GB / 2TB
Rental term: Day / Week / Month / Quarter

Node and real-time availability Use the real-time status returned by the console

Three fixed plans

Transparent configuration and term pricing

All prices are in USD. Daily terms suit defined short tasks, weekly terms fit release sprints, monthly terms support ongoing CI, and quarterly terms work well for stable build or inference services. Before choosing a longer term, confirm your migration plan for code, models, caches, and artifacts.

Light builds and validation

VMKeep M4 Core

M4 · 16GB · 256GB

$19.2/day

Weekly: $51.7/week
Monthly: $95.8/month
Quarterly: $260.6/quarter

Ideal for lightweight Xcode builds, command-line automation, dependency compatibility testing, and validating that MLX models load and respond correctly before scaling up.

Choose VMKeep M4 Core

Continuous workflows

VMKeep M4 Plus

M4 · 24GB · 512GB

$39.4/day

Weekly: $106.3/week
Monthly: $196.9/month
Quarterly: $535.6/quarter

Designed for continuously running self-hosted runners, concurrent builds for mid-sized projects, fuller dependency caches, and medium-sized MLX inference workloads that need more memory headroom.

Choose VMKeep M4 Plus

High memory and parallel tasks

VMKeep M4 Pro

M4 Pro · 64GB · 2TB

$60.8/day

Weekly: $164.1/week
Monthly: $303.8/month
Quarterly: $826.3/quarter

Ideal for high-memory MLX inference, parallel build tasks, comparing quantized versions of larger models, and workflows that retain more model files and build artifacts.

Choose VMKeep M4 Pro

Understanding term pricing

A rental term is not a discount switch for resources; it defines the order boundary for continuous use. When the end date is uncertain, start with a shorter term for validation. Once build queues, model services, or remote workstations are stable, move to a monthly or quarterly term. Before renewing, recheck the configuration, node, add-on storage, and order term.

Deployment comparison

Rental, ownership, and shared VMs have different trade-offs

There is no single answer for every use case. Renting makes it easier to control spending when you need real Apple Silicon short term, want less hardware management, or expect workloads to change with the project. Buying may suit long-term fixed use when your team can manage the equipment.

Cloud Mac rental vs. owned hardware vs. shared virtual machines
Comparison point	VMKeep Cloud Mac	Owned hardware	Shared virtual machine
Upfront cost	Pay for the selected term without taking on the full hardware purchase upfront	Requires an upfront hardware purchase plus accessories, networking, and hosting arrangements	Typically billed as a resource plan; confirm sharing and usage limits
Delivery time	After selecting the configuration, node, and term, proceed to checkout; use the real-time status returned by the console	Depends on purchasing, shipping, acceptance, networking, and remote-access setup	Usually quick to create, but it is not an independent physical node
Resource exclusivity	Every order maps to a dedicated physical machine, not a virtual machine	The buyer has exclusive use of the equipment	Underlying resources and isolation depend on the service implementation
Management responsibility	VMKeep handles node-side operation and remote recovery assistance; users handle code, data, credentials, and the application environment	The buyer handles equipment, networking, remote access, troubleshooting, and replacement	The platform manages the virtualization layer; users still manage the system and application environment
Upgrade flexibility	Choose a different configuration and term from the three-plan catalog in your next order	Upgrades usually require a new purchase, hardware replacement, or redeployment	Plans may be adjusted, but available resources and performance limits depend on platform rules
Best for	iOS/macOS builds, self-hosted runners, MLX inference, and remote graphical work	Long-term fixed workloads and self-managed internal infrastructure	General workloads that do not require dedicated physical resources

Configuration guide

Choose a plan by peak memory and task concurrency

Do not choose solely by project name. The same Xcode project can require different resources depending on dependency count, concurrent tests, caching strategy, and artifact retention. The same MLX model can also have different memory requirements based on quantization, context length, and concurrent requests.

16GB / 256GB

Light builds and inference validation

Choose VMKeep M4 Core for single-project builds, command-line scripts, dependency installation checks, small-scale testing, and validating model formats, runtime environments, and API wrappers.

Run one primary build or inference task at a time
Clean build caches and artifacts promptly by project
Record stable memory usage after the model loads

24GB / 512GB

Continuous CI and medium-sized models

Choose VMKeep M4 Plus for continuously running runners, commonly used dependency caches, mid-sized project builds, and MLX workflows that keep monitoring and logging processes alongside the service.

Plan separate space for workspaces, caches, and archives
Limit the number of build tasks launched simultaneously
Continuously monitor peak memory and disk growth

64GB / 2TB

High-memory MLX and parallel queues

Choose VMKeep M4 Pro for higher-memory models, comparing multiple quantized versions, parallel build queues, and teams that want to retain more models, caches, and build artifacts on one node.

Use separate directories for models, caches, logs, and artifacts
Limit concurrency by task type to prevent memory contention
Record time to first token, sustained throughput, and peak resource usage

Add-ons

Storage and Thunderbolt 5 interconnects priced separately

Add-ons do not change the base machine's chip or memory. Expanded storage suits workloads with growing model files, build caches, or archives. Thunderbolt 5 interconnects are priced per machine and should be selected only when the workflow requires them.

Expanded storage

+1TB SSD

Adds space for models, dependency caches, build artifacts, and temporary data. Estimate data growth over the rental term before ordering.

Day: $2.5/day
Week: $6.8/week
Month: $12.6/month
Quarter: $34.3/quarter

Expanded storage

+2TB SSD

Suitable for retaining multiple model versions, large caches, or long-term build archives. Encrypt sensitive data yourself and maintain backups.

Day: $5/day
Week: $13.6/week
Month: $25.2/month
Quarter: $68.6/quarter

Priced per machine

Thunderbolt 5 interconnect

For eligible multi-node task orchestration. It is not a general-purpose performance boost; add it only after confirming the connection method and task allocation.

Day: $1.3/day
Week: $3.5/week
Month: $6.5/month
Quarter: $17.7/quarter

VMKeep add-on pricing by rental term
Add-on	Daily	Weekly	Monthly	Quarterly
+1TB SSD	$2.5/day	$6.8/week	$12.6/month	$34.3/quarter
+2TB SSD	$5/day	$13.6/week	$25.2/month	$68.6/quarter
Thunderbolt 5 interconnect (per machine)	$1.3/day	$3.5/week	$6.5/month	$17.7/quarter

Five available nodes

All three plans cover the full node catalog

Available regions are Singapore, Japan (Tokyo), South Korea (Seoul), Hong Kong, and the US East, for five nodes or data centers. All three machine plans are generally orderable; use the real-time status returned by the console for actual availability.

Orderable matrix for five nodes and three VMKeep Cloud Mac plans
Node or data center	Region code	VMKeep M4 Core	VMKeep M4 Plus	VMKeep M4 Pro
Singapore	SG	Available	Available	Available
Japan (Tokyo)	JP	Available	Available	Available
South Korea (Seoul)	KR	Available	Available	Available
Hong Kong	HK	Available	Available	Available
US East	US-E	Available	Available	Available

How to choose a node

Prioritize the region with the most reliable connection from where your team primarily works, then test the actual path to your code repositories, artifact systems, and callers. Node selection does not replace application-level timeout, retry, and logging design.

Planning cross-region collaboration

When team members are spread across regions, first decide who handles graphical interface work and who handles automation, then have the primary operator test the SSH or VNC experience. Do not infer connection quality from the city name alone.

Pre-order checklist

Complete these six checks before creating an order

Define the task boundaries in advance to reduce repeat work caused by insufficient resources, a mismatched term, or incomplete migration preparation. Each item should have a clear owner and verifiable result.

01

Confirm the use case

Specify whether the main workload is Xcode builds, self-hosted runners, MLX inference, remote graphical work, or a combination, and estimate daily runtime.
02

Choose a node

Choose from Singapore, Japan (Tokyo), South Korea (Seoul), Hong Kong, and the US East, then have the actual operator verify the connection path.
03

Check memory and storage

Record peak memory, model size, dependency caches, build artifacts, and log growth for the current workload, then decide whether you need an SSD add-on.
04

Review the billing term

Choose a day, week, month, or quarter based on expected continuous usage. Recheck renewal before the term ends and allow time to export your data.
05

Prepare remote access

Plan SSH first for command-line and automation tasks, and VNC for graphical work. Update the password and configure SSH keys after the first connection.
06

Confirm the data migration plan

List how code, models, dependencies, caches, build artifacts, and logs will move in and out. Distribute sensitive credentials using least privilege.

Checkout details

Two payment methods, settled uniformly in USD

VMKeep supports only USDT-TRC20 and Visa / Mastercard / Amex (via Stripe). All orders are settled in US dollars (USD); the gateway available for a given order is determined by the console.

Digital asset: USDT-TRC20
Bank card: Visa / Mastercard / Amex (via Stripe)
Settlement currency: US dollars (USD)

Configure your order

Choose a configuration, node, and term, then proceed to checkout

First set memory and storage limits from the three machine plans, then choose one of five available nodes, and finally select a day, week, month, or quarter based on expected continuous use. After creating the order, view the machine, term, and management actions in the console.

1Choose a configuration
2Choose a node
3Choose a term

Configure a Cloud Mac now Read the connection guide Review security and responsibility boundaries