> See the full index of documentation pages at https://docs.zettabyte.com/llms.txt.

# The resource queue

> When the capacity you request is not immediately available, you can queue for it instead of failing.

Source: https://docs.zettabyte.com/docs/run-on-gpus/resource-queue

Last updated: 2026-07-21

When you request more resource than is free right now, zCLOUD lets you join a queue rather than erroring out. When capacity frees up and your request can be satisfied, it becomes runnable and is allocated to you.

## Joining the queue

1. When launching an instance (or creating a task) and capacity is short, the primary action changes to **Join The Queue**.
2. Select **Join The Queue** to enqueue your request for the GPU count you specified.

A queued request appears under the **Resource Queue** tab on the **Instance** page until capacity is available, with these columns: `Resource Name`, `Resource Status`, `GPU Count`, `Creation Time`, and `Unit Price`.

## How allocation works

> [!NOTE]
> **Coming soon**
> How queued requests are prioritized and granted is documented here. This section is being finalized.

## Notes and limits

- Billing for a queued request begins when capacity is allocated, not while you wait. See How billing works.
- The capacity figure shown at request time refers to available GPU capacity, *not* your wallet balance. Keep the two separate. See How billing works.