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

# Scraper Studio specifications

> Bright Data Scraper Studio infrastructure limits, billing model (CPM + per-GB downloads) and data retention windows for batch and real-time snapshots in 2024.

This reference covers Bright Data Scraper Studio's infrastructure limits, billing model, and data retention rules for batch and real-time collections.

## Billing

Bright Data Scraper Studio billing is based on page loads and file downloads.

| Item           | Unit                   | Rate                                                                  |
| -------------- | ---------------------- | --------------------------------------------------------------------- |
| Page loads     | CPM (1,000 page loads) | See [Pricing page](https://brightdata.com/pricing/web-scraper/custom) |
| File downloads | Per GB                 | Billed separately from page loads                                     |

**What counts as a page load:** One URL request processed by Bright Data Scraper Studio infrastructure, regardless of response size.

<Note>
  File downloads are billed at a separate per-GB rate and do not count toward CPM. Check your [Billing dashboard](https://brightdata.com/cp/billing/overview) for current rates.
</Note>

## Infrastructure limits

| Item                                               | Limit                         | Behavior when exceeded                                          |
| -------------------------------------------------- | ----------------------------- | --------------------------------------------------------------- |
| Parallel batch jobs                                | 100 simultaneous              | Additional jobs queue automatically                             |
| Job queue size                                     | Unlimited                     | Jobs run as capacity becomes available                          |
| `/trigger_immediate` real-time collection requests | 50K requests/min per customer | Requests above the limit return a `429 Too Many Requests` error |

## Data retention

| Data type                | Retention period | Behavior after expiry |
| ------------------------ | ---------------- | --------------------- |
| Batch collection results | 16 days          | Permanently deleted   |
| Real-time results        | 7 days           | Permanently deleted   |

<Warning>
  Export your data before the retention period expires. Bright Data does not recover expired data.
</Warning>

## Frequently asked questions

<AccordionGroup>
  <Accordion title="What happens when I exceed 100 parallel jobs?">
    Additional jobs queue automatically. No jobs are dropped or cancelled. Jobs run in the order they were submitted as capacity becomes available.
  </Accordion>

  <Accordion title="What happens to my data after the retention period?">
    Data is permanently deleted after 16 days (batch) or 7 days (real-time). Bright Data does not recover expired data. Export your results before expiry.
  </Accordion>

  <Accordion title="Are file downloads included in my CPM billing?">
    No. File downloads are billed separately at a per-GB rate. CPM covers page loads only. Both charges appear in your [Billing dashboard](https://brightdata.com/cp/billing/overview).
  </Accordion>

  <Accordion title="What counts as a page load?">
    One page load equals one URL request processed by Bright Data Scraper Studio infrastructure, regardless of page size or response content.
  </Accordion>

  <Accordion title="What is the difference between page loads and records?">
    A page load is a URL request processed by Bright Data infrastructure. In Scraper Studio, actions such as `navigate()`, `request()`, or `load_more()` can generate page loads. Page loads are used for usage and billing calculations, and are billed in CPM units, meaning per 1,000 page loads.

    A record is a structured data item extracted from the loaded page or request response. For example, if one product listing page is loaded and the scraper extracts 20 products from it, that counts as 1 page load and 20 records.

    Page loads and records are not always the same. One page load can produce zero, one, or many records, depending on the website structure and scraper logic. In Scraper Studio, billing is based on page loads by default, while records represent the amount of data returned.
  </Accordion>
</AccordionGroup>

## Related

<CardGroup cols={2}>
  <Card title="Understanding Scraper Studio" icon="book-open" href="/datasets/scraper-studio/introduction">
    Understand how Scraper Studio works and when to use it
  </Card>

  <Card title="Pricing & Billing" icon="credit-card" href="https://brightdata.com/cp/billing/overview">
    Full billing details for all Bright Data products
  </Card>
</CardGroup>