Light workloads
XOps 16
For quantized model experiments, solo development, and on-demand builds.
$19.4/day
- Chip
- M4
- Unified memory
- 16GB
- Storage
- 256GB
Day $19.4 · Week $52.3
Month $96.8 · Quarter $263.3
Dedicated physical Mac · Not a VM
Rent a Mac mini M4 in Seoul and run your workloads directly
For development teams in Northeast Asia: run MLX inference, Xcode builds, and remote macOS workflows on a cloud Mac dedicated to your use.
Choose the right fit for your memory needs, from M4 / 16GB to M4 Pro / 64GB. Pay in advance by the day, week, month, or quarter. Check live availability in the console when you place your order.
From $19.4/day · Ready in about 4 minutes
Node delivery seal
MAC / M4
One order, one dedicated physical Mac
ComputeDedicated CPU / GPU / unified memory
Remote accessSSH + VNC
Access detailsView in the console
Choose a model and billing period, then confirm the node and live availability at checkout.
Plans and billing
Choose a configuration for your models and build workloads
All prices are in USD. Switch between billing periods to compare all three plans. Storage upgrades and other add-ons can be selected at checkout.
Popular for inference
XOps 24
For running common 7B–8B quantized models and continuous builds.
$40.8/day
- Chip
- M4
- Unified memory
- 24GB
- Storage
- 512GB
Day $40.8 · Week $110.1
Month $203.8 · Quarter $554.3
Multitasking and larger models
XOps Pro 64
Extra unified memory for larger models or running multiple models in parallel.
$61.2/day
- Chip
- M4 Pro
- Unified memory
- 64GB
- Storage
- 2TB
Day $61.2 · Week $165.1
Month $305.8 · Quarter $831.8
All three plans include SSH and VNC access. Confirm available configurations, node availability, and the final amount through the live console. Pay in USD using USDT-TRC20 or Visa / Mastercard / Amex (via Stripe).
Connection guide
Connect to a Seoul node from Northeast Asia
The reference latency from Northeast Asia is 30~50ms. Use this figure as a starting point when choosing a location; it is not a fixed latency for every carrier route or connection.
Your experience also depends on your city, carrier routing, peak-time congestion, and local network. SSH command-line work is less sensitive to round-trip latency. Before using a VNC desktop or syncing large model weights, test your own connection and choose a node and transfer window accordingly.
Reference latency30~50ms
Access methodsSSH / VNC
Remote workDesktop over an SSH tunnel
Choose a location based on where your team connects from. Reference latency is not a service guarantee. Before ordering, check your own network with ping and route diagnostic tools.
Getting started
From configuration to first login in about 4 minutes
This is the typical process. Payment confirmation and actual delivery progress are shown in the console.
- 01About 1 min
Choose a plan and node, then pay
Confirm the billing period and configuration, then complete payment in USD at checkout.
- 02About 1 min
Your dedicated physical node is assigned
Once your order is confirmed, you get an entire machine with compute resources that aren't shared with other customers.
- 03About 1.5 min
System setup and access details
Delivery information and connection credentials are available in the console after you sign in.
- 04About 0.5 min
Log in with SSH or VNC
Test your connection, then install dependencies, sync your project, or start an inference service.
Before you order
Three questions about Seoul nodes
Do plans cost more at Seoul than at other locations?
The three plans and four billing periods listed here use the same USD catalog prices across locations. After choosing a node, confirm the final configuration, add-ons, and total at checkout. Availability is shown live in the console.
Can I expect 30~50ms latency on every connection from Northeast Asia?
No. The reference figure is not a guarantee for every connection. Routing, your carrier, and your local network all affect latency. If you mainly use a graphical desktop or transfer large files, test round-trip latency to the node from your work network, then use the connection guide to adjust display quality and transfer times.
Can I choose a different plan at a Seoul node after ordering?
All three listed plans are available at this node. Choose a configuration that meets your current memory and storage needs when you order. If you need to make a change later, check the available options and current configuration in the console before deciding whether to redeploy. Back up your files and model weights before migrating projects.
Ready to choose? Start with a dedicated Mac mini
Plans start at $19.4/day. Confirm the node, billing period, and live availability at checkout. Delivery typically takes about 4 minutes.