Run MLX services, Xcode builds, and remote tests on one cloud Mac, with development and inference workloads across North America West Coast in mind. The entire machine is yours, not a virtual machine. Start with a daily rental and choose the billing period that suits you.
Two configurations · Billed in USD · Check the console for live availability
All prices are listed in USD. Both configurations come with a dedicated physical machine. Select a model to preselect its configuration, then confirm the node and billing period on the order page. Additional storage and Thunderbolt 5 linking are priced separately.
M4 · 16GB unified memory · 256GB SSD
A good fit for everyday Xcode builds, device testing, and experiments with quantized 7B–14B models when you have a clear memory budget. Model weights, context, and other processes all count toward the 16GB memory limit.
M4 Pro · 64GB unified memory · 2TB SSD
For workflows that need a larger context window, multiple models resident in memory, or inference services running alongside large builds. The machine's resources aren't shared with other customers.
Optional add-ons include +1TB SSD, +2TB SSD, and Thunderbolt 5 linking. Add-ons are billed for the selected period. See all rental plans for itemized pricing.
The reference average latency to North America West Coast is 25~35 ms. Use this as a guide when choosing a location, not as a guarantee for every connection. Actual round-trip times vary with your location, provider, network congestion, and route.
If your workflow mostly involves SSH commands, CI jobs, or inference API calls over a tunnel, compare throughput and connection stability first. For extended VNC use, test responsiveness from your usual network. We recommend forwarding inference service ports over an SSH tunnel rather than exposing them directly to the public internet.
The roughly 4-minute breakdown below is a guide to the steps, not a guaranteed setup time. Actual node assignment, payment confirmation, and first connection times depend on the order process.
Pick M4 or M4 Pro based on your memory and storage needs, then choose a node and a daily, weekly, monthly, or quarterly billing period.
Confirm the configuration, add-ons, and amount in USD. Payment is accepted only via USDT-TRC20 or Visa / Mastercard / Amex processed by Stripe. Available payment gateways are determined by the response from the backend.
Check your order and instance status in the console. The console provides the live availability status; information on this page is not a substitute.
Once the instance is ready, retrieve your connection credentials in the console. Connect over SSH first, then use the VNC interface if your workflow requires it.
Check that the node and machine fit your workflow before checkout. The answers below don't replace the live configuration shown on the order page.
On the order page, check the node name in the node selection section, then review the order summary again before paying. A model button only preselects the model; it doesn't preselect a node for you.
No. 25~35 ms is a reference average latency for the area. Your network provider and connection route also affect the result. If you plan to use the graphical interface regularly, judge performance from your own network connection.
Both models are listed for this node. Choose based on model memory, project size, and storage needs, then confirm the node and billing period on the order page. Check the console for live availability.
Choose a model, billing period, and node to see the amount due. Review add-ons before paying. If you're unsure about memory requirements, compare the full plans first.