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

# Compute

> Provision and manage cloud servers for your Kubernetes clusters

<Frame>
  <img className="block dark:hidden" src="https://mintcdn.com/akua-1dce587a/KB6u5PJRNHAXrgro/images/heros/compute-overview-light.svg?fit=max&auto=format&n=KB6u5PJRNHAXrgro&q=85&s=0b62a12088c5ea14ec6fbb10522bb833" alt="Account configs — Akua-managed and your own cloud key (BYOM) — feed machines into one cluster, with each machine tagged by which account owns it" width="1536" height="864" data-path="images/heros/compute-overview-light.svg" />

  <img className="hidden dark:block" src="https://mintcdn.com/akua-1dce587a/KB6u5PJRNHAXrgro/images/heros/compute-overview-dark.svg?fit=max&auto=format&n=KB6u5PJRNHAXrgro&q=85&s=950234596ec17b6aefec796891bb7a9b" alt="Account configs — Akua-managed and your own cloud key (BYOM) — feed machines into one cluster, with each machine tagged by which account owns it" width="1536" height="864" data-path="images/heros/compute-overview-dark.svg" />
</Frame>

Akua provisions and manages compute resources: the cloud servers that run your workloads. Managed compute and BYOM provisioning use the providers available in your dashboard; Hetzner Cloud is available today. You don't need a separate cloud provider account to get started, and you can still attach existing workers when that better fits your setup.

## How it works

Akua scales your compute automatically. When your workloads need more resources, Akua detects unschedulable pods and provisions a new server. When demand drops, idle servers are removed to save costs.

```
Pod needs resources → Akua provisions a server → Server joins the cluster → Pod scheduled
```

The entire flow takes about 2 minutes from pending pod to running workload.

Akua keeps the setup simple. Your workspace plan and cluster settings determine which regions, credentials, and instance types are available, so you only choose what you need.

## Three compute paths

### Akua-managed machines

Akua provisions servers for you in Hetzner Cloud. No cloud credentials are required. Machines are subject to [plan quotas](/platform/quotas) and [billing](/billing/overview).

| Plan | Machines | CPU     | Memory | Instance types | Machine lifetime |
| ---- | -------- | ------- | ------ | -------------- | ---------------- |
| Free | 1        | 2 cores | 4 GB   | Small only     | 7 days           |
| Pro  | 3        | 8 cores | 16 GB  | All available  | No limit         |

### Bring your own cloud key

Add your own supported cloud provider key and Akua provisions machines in your account. You pay the provider directly, with no machine quotas, no lifetime limits, and access to any instance type your account supports.

<Steps>
  <Step title="Add your API key">
    Go to **Compute** in the dashboard and click **Add Credential**. Akua validates the token before saving it.
  </Step>

  <Step title="Create a compute config">
    Click **Create Config**, select your credential, choose a region, and name the configuration.
  </Step>

  <Step title="Provision machines">
    Akua uses your config to provision machines in your cloud account when your workloads need resources.
  </Step>
</Steps>

Akua stores the credential securely and re-validates it periodically. You'll get a dashboard notification if the credential becomes invalid.

### Manually attached workers

Attach an existing server to a managed cluster when you want to keep full ownership of the machine without using managed machine provisioning. Use the worker bootstrap command or the cloud-init path in [Add workers](/clusters/add-workers). Imported clusters keep worker management with their provider or cluster administrator.

## Provisioning limits

Akua enforces limits at two levels to keep your infrastructure predictable:

* **Allocation quota**: how many machines your workspace can own (free: 1, pro: 3). BYOM machines are exempt.
* **Concurrency limit**: how many machines can be launching at once per cluster (free: 1, pro: 3). This prevents runaway provisioning loops from multiplying in-progress requests.

Both limits are enforced server-side and take effect immediately when your plan changes, with no cluster reconfiguration needed.

See [quotas](/platform/quotas) for the full table of limits per plan.

## Supported providers

<CardGroup cols={2}>
  <Card title="Hetzner Cloud" icon="cloud" href="/compute/providers/hetzner">
    Current managed compute and BYOM provider details.
  </Card>

  <Card title="More coming" icon="plus">
    AWS, GCP, Azure, DigitalOcean, Vultr (planned).
  </Card>
</CardGroup>

## What's next?

<CardGroup cols={2}>
  <Card title="Machines" icon="server" href="/compute/machines">
    Lifecycle, provisioning, deletion, and auto-scaling controls.
  </Card>

  <Card title="Hetzner Cloud" icon="cloud" href="/compute/providers/hetzner">
    Regions, instance types, and pricing.
  </Card>

  <Card title="Quotas" icon="gauge" href="/platform/quotas">
    Understand resource limits per plan.
  </Card>
</CardGroup>

## Related topics

<CardGroup cols={2}>
  <Card title="Add workers manually" icon="terminal" href="/clusters/add-workers">
    Use a bootstrap command to attach your own servers.
  </Card>

  <Card title="Clusters" icon="server" href="/clusters">
    Understand how clusters work and how compute fits in.
  </Card>

  <Card title="Plans and pricing" icon="credit-card" href="/pricing">
    Compare plans and understand machine limits.
  </Card>

  <Card title="Compute API" icon="code" href="/api-reference/machines/list-machines">
    Manage machines programmatically.
  </Card>
</CardGroup>
