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

# Compute Units

## What Are Compute Units?

**Compute Units (CUs)** represent the amount of computational resources consumed by your API calls. They provide a standardized way to measure and compare usage across different types of requests.

## Why Use Compute Units?

Compute Units enable transparent and fair pricing. Instead of paying flat rates, you’re billed based on the actual computational cost of your requests.\
This approach ensures:

* You don’t overpay for lightweight API calls.
* Complex operations are accurately billed according to the resources they consume.

## Throughput

Some services, such as Paymaster and Bundler, include a throughput limit (e.g., requests per minute). If throughput autoscale is enabled, requests that exceed this limit are still processed - billed using additional Compute Units.

## Autoscale Compute Units

When you reach the Compute Units included in your plan, **autoscale** ensures uninterrupted usage. Any additional consumption beyond your quota is billed at the end of the month.

### Paymaster & Bundler

These products support two types of autoscaling:

1. **Monthly Autoscale** – Allows usage beyond your monthly CU quota.
2. **Throughput Autoscale** – Automatically scales during high-traffic spikes.

### Private RPCs

Supports **monthly autoscale**, enabling continued usage even after your quota is reached.

### Functions

* Does **not support autoscaling**.
* Usage is capped at your plan’s limit.

<Note>
  You can manage all autoscaling settings per product directly in your dashboard.
</Note>