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

# Web Search

ASSIST AI can browse the internet to answer questions requiring current or specialized information the AI may not be confident about. Users can toggle the **Web Search** Action on or off at any time to override the AI's default behavior.

When **Internal Search** is also used, ASSIST AI seamlessly combines web sources with internal documents — both in the generated answer and in the UI.

## Requirements

Web Search relies on two components: **1) a search provider** to retrieve results, and **2) a web scraper** to extract full page content for more complete answers.

| Name       | Provider                        | Price                                  | Benefit                                                     |
| :--------- | :------------------------------ | :------------------------------------- | :---------------------------------------------------------- |
| Exa        | [Exa.ai](http://Exa.ai)         | \$5 / 1,000 queries                    | AI-optimized with strong semantic search                    |
| Serper     | [Serper.dev](http://Serper.dev) | Free tier + \$0.30 / 1,000 queries     | Fast and budget-friendly                                    |
| Brave      | Brave Search API                | Usage-based                            | Independent index, wide web coverage, simple setup          |
| Google PSE | Google                          | Free tier + \$5 / 1,000 queries        | Official Google results, ideal for security-conscious teams |
| SearXNG    | SearXNG                         | Self-hosted / infrastructure cost only | Private, self-hostable metasearch across multiple engines   |

## Web Scrappers

Since search providers return only short snippets, a scraper fetches the full page content for richer, more reliable responses.

| Name                  | Requires Config          | Benefits                                                                                        |
| :-------------------- | :----------------------- | :---------------------------------------------------------------------------------------------- |
| ASSIST AI Web Crawler | No                       | Built-in, works on most pages, free, and keeps queries within your deployment                   |
| Firecrawl             | Yes — API key + base URL | Higher performance with better handling of edge cases                                           |
| Exa                   | Yes — API key            | Fetches and summarizes page content, ideal for teams already using Exa as their search provider |