Guides
The reference answers "what does this endpoint take". These pages answer "how do I get this done", in the order you would actually do it. Every call in them was run against a live project before it was published.
Getting started
Issue a token, make your first call, create your first server, and point Terraform or OpenTofu at the same project. Start here if you have never called this API.
A server from start to finish
One continuous run: a private network, a server with a public address, the firewall, a backup, and a clean-up that leaves nothing billable behind. Read this before you write a script that creates and deletes servers on its own.
Reseller: one project per customer
An account key, a project for each of your customers, labels to find them, keys you can hand out, costs per customer, and closing a project when a customer leaves.
Billing data
Balance, debt and the day non-payment would suspend services, invoices with their lines and PDFs, receipts, payments and balance movements, a forecast of the month, and the price of a server before you create it - read from a program with an account key that has billing access.
Server console
A one-time ticket for the screen and keyboard of a running server: the ready console page, your own noVNC client, and how a portal of your own proxies the console WebSocket to its customers.
Webhooks and budgets
Invoices, the balance, project changes and budget alerts pushed to your own HTTPS address, signed per Standard Webhooks; a budget per project that alerts as its charges grow.
GPU servers
How a packaged GPU server is ordered, what the package decides for you, which fields are refused on it, and what the capacity refusal means.
GPU Pro
The GPU type where you pick the cards yourself: how to get access to it, how to size it, how to choose the host it lands on, and what the capacity refusal is telling you. Opened per project on request.