Rent a dedicated physical Mac mini, choose a location and billing period, and get your connection details in the console after payment. Follow the steps below until you can run your first xcodebuild command or start an MLX service.
The 4-minute estimate illustrates the process; it is not a delivery-time guarantee. Check real-time availability in the console.
Deciding on your workload, memory and location upfront is easier than changing your setup after payment. Check each item; no client installation is required.
These times are estimates to help you plan; individual orders may take longer or less time. Check payment and machine allocation status in the console.
Choose VMR Lab M4 or VMR Lab Pro on the setup page, then select a location and daily, weekly, monthly or quarterly billing. Only options available for the selected model are shown. Check the SSD add-on and total price before placing your order.
Choose USDT-TRC20 at checkout, or pay with Visa, Mastercard or Amex via Stripe. After payment, check your order status; closing the payment page doesn’t confirm that your order is complete.
The console shows your order and instance status. Wait for machine allocation and system readiness; don’t place another order before the status updates. If the status needs checking, sign in to the console and submit a support ticket with your order number.
Find your SSH connection details and VNC instructions in the instance details. First verify your login and command line over SSH, then open the graphical desktop if needed. Don’t share connection credentials in public repositories or chat logs.
About 4 minutes in total. Network conditions, payment confirmation and instance readiness can affect the actual time. If something goes wrong, start troubleshooting with the console records.
These examples show SSH, an Xcode build, a TestFlight upload and an MLX service. The logs illustrate output format; they are not performance results from a real task. Get your connection details in the console, then set the required environment variables in the commands.
ssh -i ~/.ssh/vmr_lab "$VMR_USER@$VMR_HOST"After connecting, your terminal will show the remote macOS command prompt.
xcodebuild -project "$PROJECT" -scheme "$SCHEME" build
fastlane betaOn success, you’ll see ** BUILD SUCCEEDED **; fastlane will report the upload result.
python3 -m pip install mlx-lm
python3 -m mlx_lm.server --model "$MODEL_PATH" --host 127.0.0.1 --port 8080After starting the service, check its logs, then access port 8080 locally through an SSH tunnel.
Before you start: VMR_USER and VMR_HOST should match the instance details; PROJECT, SCHEME and MODEL_PATH should point to your project or model directory. fastlane beta requires you to configure your own release pipeline in advance. Don’t bind the inference service directly to a public address.
Set up a reliable, revocable access path first. That way, you won’t have to work out access again when automating builds or sharing a model with your team.
Generate a key pair locally or choose a dedicated one, then add the public key to your instance as instructed in the console. Keep the private key on a trusted device. Use a separate key for CI and document its purpose so you can revoke it later.
Verify the host fingerprint on first connection. If multiple collaborators need access, don’t share the same private key among team members.
For Xcode, the Simulator or video workspaces, follow the VNC instructions in your instance details. Keep command-line builds in SSH so you don’t need to maintain a graphical session just to check logs.
Remote desktop is available after you sign in to the instance. Get connection details from the console and don’t put credentials in public documentation.
Run mlx_lm.server on the machine’s local address, 127.0.0.1:8080, then access it with SSH local port forwarding. This lets you debug the API from your own computer without exposing the service directly to the public internet.
ssh -i ~/.ssh/vmr_lab -L 8080:127.0.0.1:8080 "$VMR_USER@$VMR_HOST"For help with installation, connections or billing periods, start at the Help Center. To decide how to use a machine for builds, inference or audio and video tasks, explore the relevant workflows.
Choose a model, location and billing period. Final configuration and availability are shown in real time in the console. Want to compare prices first? Explore the two plans.