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

# Pages quota

> How SiteGPT measures training content, and what to do when you hit the limit.

Everything your chatbot learns is measured in **pages**. One page equals
**2,500 cleaned characters**, roughly 400 to 500 words after navigation,
markup, and boilerplate are stripped.

The same measure applies to every content type: a crawled web page, an
uploaded file, a synced Notion doc, or a text snippet. A long document counts
as several pages; a short one may count as one.

## Quota by plan

| Plan       | Pages   |
| ---------- | ------- |
| Starter    | 1,000   |
| Growth     | 10,000  |
| Scale      | 50,000  |
| Enterprise | 500,000 |

The quota is shared across all chatbots on your account. Check your current
usage on the **Usage** page (top navigation).

## What counts, what does not

* Text content counts after cleaning. Images and video on a page do not
  count; only the page's text does.
* YouTube videos count by their transcript length.
* Deleting links or files frees their pages immediately.
* Resyncing existing content does not consume additional quota; the page
  count adjusts to the content's current length.

## When you hit the limit

Imports and crawls stop when the quota is used up, and affected documents can
show a quota-blocked failure status.

1. **Trim what you do not need.** Sort your Links List by what your visitors
   never ask about; marketing pages, old blog posts, and event pages are
   common dead weight. Deleting them frees pages instantly.
2. **Target the crawl.** Use **Max Pages to Scrape**, include/exclude
   selectors, and sitemap imports instead of unbounded crawls. See
   [scrape options](/docs/content/website-links#scrape-options).
3. **Upgrade.** Compare quotas on the [pricing page](https://sitegpt.ai/pricing),
   or use the **Upgrade** prompts in the dashboard.