> ## Documentation Index
> Fetch the complete documentation index at: https://docs.nodepick.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Compute Nodes: Isolated Linux Environments on Demand

> Nodepick nodes are isolated, dedicated Linux kernel environments with full SSH and root access, provisioned in seconds and billed per second.

A Nodepick node is a fully isolated, dedicated Linux kernel environment that you can provision in under 30 seconds. Each node gives you complete control over your compute environment — from the kernel up — with no shared neighbors, no noisy resources, and no configuration guesswork. Whether you're running a long-lived service, a short burst workload, or an automated pipeline, nodes give you a clean, root-accessible machine exactly when you need it.

## Node Anatomy

Every node you provision on Nodepick is built from the same set of core primitives. Understanding what each component does helps you choose the right configuration for your workload.

* **Dedicated Linux kernel** — Each node runs its own isolated kernel. There is no hypervisor overhead and no shared kernel surface with other tenants.
* **Full root access** — You get unrestricted `root` access to the environment. Install packages, configure services, modify kernel parameters, or run privileged processes without restriction.
* **SSH connectivity** — Every node exposes an SSH endpoint from the moment it reaches the `running` state. You connect directly using standard SSH tooling.
* **Routable IP address (optional)** — You can attach a routable public IP to your node so it is reachable from the internet. If your workload is internal-only, you can skip this to reduce cost.

## Resource Configuration

When you create a node, you select the resources you need. All resources are billed per second from the moment the node is created until the moment it is deleted.

| Resource    | Unit        | Price per Second |
| ----------- | ----------- | ---------------- |
| CPU         | per core    | \$0.000014       |
| RAM         | per GB      | \$0.0000045      |
| Storage     | per GB      | \$0.00000003     |
| Routable IP | per address | \$0.00000076     |

You can combine these resources in any configuration that fits your workload. A small task might need a single core and 1 GB of RAM, while a compute-heavy job might use many cores and significantly more memory.

## Node Lifecycle

Every node moves through a simple, predictable lifecycle. Understanding each state helps you manage costs and plan your workloads.

<Steps>
  <Step title="Created">
    You request a node via the dashboard, SDK, or API. Nodepick begins provisioning immediately. The node enters a `provisioning` state and is typically ready in under 30 seconds. **Billing starts at this moment.**
  </Step>

  <Step title="Running">
    The node is fully provisioned and reachable. You can connect via SSH, deploy workloads, and use all attached resources. The node stays in this state until you explicitly delete it.
  </Step>

  <Step title="Deleted">
    You send a delete request. The node is permanently removed and all resources are released. **Billing stops immediately.** Deleted nodes cannot be recovered.
  </Step>
</Steps>

<Note>
  Billing begins the moment a node is created and stops the moment it is deleted. There are no minimum usage windows — you are charged only for the exact seconds your node exists.
</Note>

## Explore the Nodes Documentation

<CardGroup cols={2}>
  <Card title="Create a Node" icon="circle-plus" href="/nodes/create">
    Provision a new node in under 30 seconds using the dashboard, Python SDK, or HTTP API.
  </Card>

  <Card title="Manage Nodes" icon="sliders" href="/nodes/manage">
    List, inspect, and delete your nodes to monitor status and keep costs under control.
  </Card>

  <Card title="SSH Access" icon="terminal" href="/nodes/ssh-access">
    Connect to your node with full root access using standard SSH tooling.
  </Card>

  <Card title="Pricing" icon="receipt" href="/nodes/overview#resource-configuration">
    Understand per-second billing across CPU, RAM, storage, and IP addresses.
  </Card>
</CardGroup>