Cloud Mac Dedicated Physical Server Terms of Service

VMMini grants time-limited use of a dedicated Mac mini M4 physical server. The current catalog configuration is VMMini M4, with an M4 chip, 16GB RAM, and a 256GB SSD. Each valid order corresponds to resources on a designated physical node. It is not a virtual machine, and the host computing resources assigned to the order are not shared with other customers.

The service is intended for MLX inference, Mac AI deployment, iOS or macOS continuous integration, remote automation, and other lawful development work. The service includes use of the physical node listed in the order, required connection details, order management, and technical support within the agreed scope. It does not include development or maintenance of the user’s code, models, build scripts, or third-party software.

Before creating an order, completing payment, or starting to use a node, the user should review these Terms, the order page, and applicable policies. Creating an order means the user has read and accepted the Terms effective at that time and confirms they are authorized to place the order for themselves or their organization.

The user must provide accurate, usable contact information that they are authorized to use and ensure that the email address receiving order notifications can receive messages. Actions performed on the node by authorized team members, automation, or credential holders are deemed to be within the user’s authorization.

Orders may be created by the day, week, month, or quarter. Each order must clearly record the model, node, billing cycle, selected storage add-ons, number of Thunderbolt 5 links, start time, expiration time, and total amount due. These records are governed by the order details shown in the console.

The current catalog nodes are Singapore, Japan (Tokyo), South Korea (Seoul), and Hong Kong. VMMini M4 is an available catalog combination in all four locations. Actual availability when an order is created is determined by the real-time result returned by the console. Cities or configurations not listed in the catalog are not available for ordering.

An order enters fulfillment after payment is completed and confirmed by the system. The start and expiration times are determined by the order record, not recalculated from the user’s first connection, first task run, or first data download. Before ordering, the user should select a term that matches the task duration to avoid builds, inference, or data exports extending past expiration.

Renewal is a new billing confirmation. Before the current order expires, the user should verify the next term, node, configuration, and add-ons. Unless the order page explicitly displays and the user confirms an extension, the service will not automatically continue solely because the user retains connection settings.

Node replacement is not included by default in the original order. If the user needs to change nodes, they must first submit a ticket through the console. The service team will determine the handling method based on the target node, remaining term, data migration scope, and actual availability. Users must not use the original node’s connection details directly on another node.

All prices are listed and settled in US dollars (USD). The current VMMini M4 base configuration costs $20.5/day, $55.4/week, $102.6/month, or $279.1/quarter. The order total consists of the base term, selected add-ons, and any other applicable amounts shown on the order page.

Payment methods are limited to USDT-TRC20 and Visa / Mastercard / Amex (via Stripe). The gateway actually available is determined by the backend response when the order is created. Users must verify the payment network, amount, currency, and order ID. Users are responsible for checking any delays caused by selecting the wrong network or sending payment to an address not specified for the order.

Applicable fees, taxes, refund conditions, and payment status are governed by the order page and applicable policies. Any refund or amount adjustment must correspond to verifiable order, payment, and service records and cannot be replaced by an oral statement not recorded in the order.

Users may use the node for software development, testing, builds, MLX inference, Mac AI deployment, continuous integration, and automation within lawful limits. Users must ensure that their code, models, data, dependencies, licenses, and task outputs have lawful origins and comply with applicable intellectual property, data protection, and export control requirements.

Users should apply least-privilege access management and securely protect SSH private keys, console credentials, automation tokens, and project keys. Sensitive credentials should be used through environment variables or controlled secret storage and must not be written to public repositories, build logs, image layers, or scripts accessible to unauthorized persons.

Users must not use the node for unauthorized access, credential harvesting, malicious code distribution, network disruption, traffic amplification, bypassing access controls, or interference with other customers or infrastructure. Continuous high-risk scanning, abnormal traffic, or malicious tasks must not affect the node network, support systems, or other services.

Users must not create orders by impersonating another person, resell node access without authorization, or bypass the order scope by sharing connection details. If credentials are exposed or there is an unusual login, unknown process, or unexpected network connection, the user must immediately revoke the relevant credentials, preserve necessary logs, and submit a ticket through the console.

Unavailability caused by node hardware, host network, or delivery system failures within VMMini’s control, and verifiable through service records, is eligible for an availability review. Users must provide the order ID, affected node, occurrence time, duration, error details, and necessary redacted logs.

Unavailability caused by the user’s software configuration, dependency conflicts, terminated processes, a disk filled by tasks, incorrect access rules, expired keys, code defects, or the user’s own shutdown does not count as node service unavailability. Problems with the user’s local network, code source, third-party dependencies, carrier routes, or other external systems are also not node failures directly controlled by VMMini.

Force majeure, public network outages, attacks by users or third parties, measures lawfully taken by public authorities, and other events beyond reasonable control will be assessed based on the records and actual impact. The service target does not guarantee specific latency, model throughput, build speed, or the continuous availability of third-party software.

Compensation requests must be submitted through a console ticket and supported by order records, monitoring records, the incident timeline, and the review outcome. The applicable form and amount are determined by the result displayed after review and applicable policies; a single client-side speed test or an unverifiable subjective experience is not sufficient on its own.

When the system detects a clear security risk, credential exposure, unusual network activity, use that violates these Terms, a legally required measure, or another event that may continue to affect service security, VMMini may restrict relevant connections, suspend tasks, or suspend node access. The scope of action must match the risk, and necessary order and security records will be retained.

The service may terminate when an order expires, the user initiates termination, payment is incomplete, or a serious violation is verified. Before expiration, the user must export code, models, build artifacts, logs, and other necessary data, and revoke runners, deployment keys, and automation tokens. Retaining data or connection access after expiration is not part of the default order.

If the user believes a suspension or termination was made in error, they should submit the order ID, node, time of occurrence, and relevant evidence through the console. VMMini will verify the matter against payment status, access records, security logs, and ticket information. To protect other customers and infrastructure, necessary access restrictions may remain in place until the risk is confirmed.

To the extent permitted by applicable law, each party is responsible for its own conduct, personnel, credentials, code, data, and third-party services it uses. Liability must be assessed in light of direct causation, verifiable records, the order amount, actual losses, and applicable legal requirements; indirect assumptions or unverifiable expected benefits may not serve as the sole basis.

Updates to these Terms will be published through in-product announcements, the Help Center, or console records. Updates that materially affect the service scope, billing, or user rights apply to subsequent orders or renewal confirmations from their published effective date. Orders created before an update will be handled under the records and policies applicable when they were created, unless otherwise required by law.

These Terms are governed by the laws of the jurisdiction where the platform operator is based. For disputes arising from these Terms or the service, the parties should first negotiate based on order, payment, monitoring, and ticket records. If negotiation does not resolve the dispute, it may be submitted to a court of competent jurisdiction in that jurisdiction.