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

# Choosing a method

> Learn about the different types of website lookups available

Lookups allow you to retrieve and integrate external information into your
documentation workflow.

## Website Lookup Methods

There are three primary ways to integrate website data. Choosing the right
method depends on your specific needs regarding data freshness, reliability, and
the target website's structure.

### 1. Live Web Search

This method performs a real-time search via a search engine. You can set it up to
include domains, which narrows the results to a specific website, or you can
exclude domains to remove certain websites from the search.

* **Best for:**
  * Large websites with many pages that can't be fully imported.
  * Searching for general information across the internet.
* **Limitations:** It relies on the search engine's indexing and search
  algorithms.

### 2. Import website

This method imports and indexes pages from a website, optionally including
subpages, directly into the Typewise knowledge base. It features an advanced
search algorithm that can find information more reliably and faster.

* **Best for:** Static websites, FAQs or knowledge bases.
* **Limitations:** This method has a capacity limit of 100 pages with a maximum
  depth of 10 subpages, after which it stops looking for new pages. It is **not
  suitable** for indexing entire large sites with thousands of pages.

<Note>In the platform UI, Import website appears under the **Stored knowledge** category.</Note>

### 3. Live URL fetch

This method fetches the latest content from any URL in real time. It supports
dynamically generated URLs, e.g. to get an order status. The AI Agent decides
how to fill the dynamic parameters automatically based on the context.

* **How it works:** You provide a URL template with placeholders.
  * *Example 1:*
    `https://www.hotel-search.com/search?minPrice={minimum_price}&maxPrice={maximum_price}`
    (Agent fills in price range)
  * *Example 2:* `https://www.online-shop.com/en/search?q={product}` (Agent
    searches for a specific product)
  * *Example 3:* `https://www.search-engine.com/en/search/?q={search_query}`
    (Agent inputs general search terms)
* **Best for:** Sites with predictable URL structures where you need to pass
  dynamic values like order IDs, tracking numbers, or search terms.
* **Limitations:** This method will only analyze a single page of the website.
  It won't follow links to other pages.
* **Recommendation:** Only use this method if you need dynamic parameters or the
  latest information. If the URL is static, **Import website** is usually a
  better choice.

## See also

* [Live web search](/documentation/knowledge/web/live-web-search): Search the
  web in real time
* [Import website](/documentation/knowledge/web/import-website): Import and
  index website content
* [Live URL fetch](/documentation/knowledge/web/live-url-fetch): Fetch the
  latest content from any URL in real time