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

# Usage Analytics

> Track bandwidth, requests, and destinations for your proxy traffic, and spot problems before they cost you bandwidth.

The requests dashboard gives you a detailed breakdown of how your proxies are being used. Find it at [flameproxies.com/dashboard/requests](https://flameproxies.com/dashboard/requests).

## What you can see

The dashboard shows precise information about your proxy traffic, including:

| Metric | What it tells you |
| - | - |
| **Bandwidth** | How much data your traffic transferred. This is what your balance is spent on. |
| **Requests** | How many requests went through the proxies. |
| **Host destinations** | Which target hosts your traffic went to. |
| **Average bytes** | The average size of a request, which shows how heavy the pages you load are. |
| **Timestamp** | When the traffic happened, so you can match it to your own logs and jobs. |

## Checking usage with the API

You can also read usage programmatically with the [Customer API](/api/overview):

* [`GET /api/customer/packages/:id`](/api/packages/get) returns live usage for any package you own.
* [`GET /api/customer/non-api-packages/usage`](/api/non-api-packages/usage) returns live usage for all your regular (dashboard-purchased) packages at once.

Usage comes back in bytes and GB:

```json theme={null}
{
  "bytes_used": 1200000000,
  "max_bytes": 5000000000,
  "remaining_bytes": 3800000000,
  "gb_used": 1.2,
  "gb_total": 5,
  "gb_remaining": 3.8,
  "percent_used": 24
}
```

## How to spot issues

A few patterns worth knowing:

**Bandwidth is higher than expected.** Check **host destinations** and **average bytes**. Large responses — images, video, fonts, and scripts — use much more bandwidth than the HTML or JSON you actually need. If you use a headless browser, block resources you don't need.

**One host takes most of the traffic.** Look for a retry loop or a job that ran longer than expected against that host.

**Average bytes suddenly jumps.** The target may be serving heavier pages than before, or returning error pages or CAPTCHA pages instead of data. Check a sample response.

**No traffic shows up.** Check that your requests authenticate with the right username and password, and that you're using the credentials of the package you're looking at.

<Tip>
  Give each workload its own [sub-user](/dashboard/sub-users). Separate credentials make it much easier to see which job is using your bandwidth.
</Tip>

## Next steps

<CardGroup cols={2}>
  <Card title="Sub-users" icon="users" href="/dashboard/sub-users">
    Split bandwidth between teams and workloads.
  </Card>

  <Card title="Billing & top-ups" icon="receipt" href="/dashboard/billing">
    Pricing, payment methods, and invoices.
  </Card>
</CardGroup>