Tavily

AI-oriented web search, page extraction, site crawling and site mapping.

5 actions API key 5 read0 write Calls api.tavily.com

Connect

TypeScript

const conn = await setu.connections.initiate({ userId: "user_1", toolkit: "tavily" });
console.log(conn.connectUrl); // show this to your user
await conn.waitUntilActive();

Permissions requested

This toolkit uses a single credential and requests no scopes.

Actions

Crawl a websitereadtavily.crawlCrawl a website from a root URL and return the extracted content of each page found. Can take a while and costs credits per page.

Input	Type		Description
`url`	string	Required	Root URL to begin the crawl
`instructions`	string	Optional	Natural-language instructions for the crawler
`max_depth`	integer	Optional	How far from the root to go, 1 to 5 (default 1)
`max_breadth`	integer	Optional	Links followed per page, 1 to 500 (default 20)
`limit`	integer	Optional	Total pages to process (default 50)
`select_paths`	array of string	Optional	Regex patterns of paths to include
`exclude_paths`	array of string	Optional	Regex patterns of paths to exclude
`format`	string	Optional	Output format

TypeScript

await setu.execute({
  userId: "user_1",
  action: "tavily.crawl",
  input: {
    "url": "<url>"
  },
});

Extract page contentreadtavily.extractFetch and extract the readable content of one or more URLs (up to 20) as markdown or text, optionally reranked by a query.

Input	Type		Description
`urls`	array of string	Required	URLs to extract, 1 to 20
`query`	string	Optional	Rerank extracted chunks by relevance to this query
`extract_depth`	string	Optional	basic or advanced (tables, embedded content)
`format`	string	Optional	Output format

TypeScript

await setu.execute({
  userId: "user_1",
  action: "tavily.extract",
  input: {
    "urls": []
  },
});

Get API usagereadtavily.get_usageGet this API key's credit usage and limit, plus the account's plan and usage.

No input.

TypeScript

await setu.execute({
  userId: "user_1",
  action: "tavily.get_usage",
  input: {},
});

Map a websitereadtavily.mapBuild a map of a website: return the list of URLs found from a root URL without extracting page content.

Input	Type		Description
`url`	string	Required	Root URL to begin mapping
`instructions`	string	Optional	Natural-language instructions for the crawler
`max_depth`	integer	Optional	How far from the root to go, 1 to 5 (default 1)
`max_breadth`	integer	Optional	Links followed per page, 1 to 500 (default 20)
`limit`	integer	Optional	Total links to process (default 50)
`select_paths`	array of string	Optional	Regex patterns of paths to include
`exclude_paths`	array of string	Optional	Regex patterns of paths to exclude

TypeScript

await setu.execute({
  userId: "user_1",
  action: "tavily.map",
  input: {
    "url": "<url>"
  },
});

Search the webreadtavily.searchSearch the web with Tavily and return ranked results with content snippets, optionally with a short LLM-written answer.

Input	Type		Description
`query`	string	Required	The search query
`search_depth`	string	Optional	Search depth
`topic`	string	Optional	Search topic
`max_results`	integer	Optional	Number of results, 0 to 20 (default 10)
`time_range`	string	Optional	Only results from this period
`include_answer`	boolean	Optional	Include an LLM-generated answer
`include_domains`	array of string	Optional	Only search these domains
`exclude_domains`	array of string	Optional	Never return these domains

TypeScript

await setu.execute({
  userId: "user_1",
  action: "tavily.search",
  input: {
    "query": "<query>"
  },
});