How does usage-based billing work?+
Projects pay for resources they create and the time those resources run. Duration-based resources are billed per minute, capped at 672 hours per monthly invoice window for the same resource slot. The console shows the exact regional plan total before you launch.
When does billing start and stop?+
Compute billing begins when provisioning starts and ends when deletion completes. Provisioned or retained storage bills until deleted. Stopping a virtual machine does not remove the charges for its retained volumes.
How is AI inference billed?+
All available AI inference is paid from the first token, including open and Azure Foundry models. Usage is metered at each model’s listed input and output rates. Premium Azure Foundry token rates are now 50% off LayerRail’s standard rates, automatically applied to new usage. Cloudflare models use Cloudflare’s published price plus 10%. There is no free monthly inference token allowance.
Which models receive the 50% premium discount?+
All premium Azure Foundry models receive 50% off LayerRail’s standard token rates. The console shows the discounted prices, and the discount applies automatically to new usage without a coupon. Listed examples are base rates; models with long-context pricing use their listed long-context tier, also at 50% off. Model availability is shown in the console.
When is a separate AI invoice issued?+
AI usage is separate from subscriptions and other cloud charges. An invoice is issued at $10 accrued, or at month end for a balance of at least $1. Smaller balances carry forward. Existing monetary credits apply before payment. Unpriced models are unavailable.
How do I pay a usage invoice?+
Follow the Bachs checkout link for the invoice on your project billing page. Further inference pauses when payment is due and resumes when payment is confirmed. Checkout is required to pay; this flow does not automatically charge a saved card.
What is included with a virtual machine?+
CPU and memory are included in the selected plan. Root storage and public IPv4 are included where marked. GPU capacity is metered when you choose a GPU plan. Check the console for the exact plan and regional total before launching.
How are GitHub runners billed?+
Runner minutes are recorded after a job starts, from $0.001 per minute for standard-2. Runner cache storage is $0.04 per GB-month and extra concurrency is $1 per month. A project receives a $1 monthly runner credit when it has runner usage.
Do Game VPS plans include additional usage charges?+
Game VPS packages are fixed monthly packages from $3 to $75. Each includes a 128 GB Windows system SSD. Higher packages add CPU and memory. Checkout shows the exact payment total before provisioning.
Is there a discounted annual rate?+
The currently published Game VPS monthly and annual-view rates are the same. Usage-based resources stay metered, with no annual discount advertised. Contact us for reserved capacity and dedicated requirements.
Can I self-host LayerRail?+
Yes. The control plane is open source under AGPL-3.0. Run the console, workers, and database on your own infrastructure with your own provider account. Your underlying infrastructure and provider bills remain your responsibility.
Which regions are available?+
The published region list includes East US, East US 2, Central US, UK South, and South Africa North. Availability differs by service. The regions guide and console show current support.
What happens if an invoice is unpaid?+
Outstanding invoices remain visible until paid. Services can be restricted, suspended, or eventually deleted according to the billing terms. Contact support@layerrail.com for invoices, credits, or payment questions.
