cotera
Logo

Tavily

Authentication Type: API Key Description: Search the web and extract content from URLs using Tavily, a search engine and web scraping solution optimized for LLMs.

Authentication

To authenticate, you'll need a Tavily API key. Sign up at tavily.com to get 1,000 free searches, then find your API key in the Tavily dashboard.

Search the web using Tavily Search, returning relevant results with content snippets and optional AI-generated answers.

Operation Type: Query (Read)

Parameters:

  • query string (required): The search query to execute
  • searchDepth string: Controls latency vs relevance. Options: "basic" (default, balanced), "advanced" (highest relevance, more latency), "fast" (lower latency), "ultra-fast" (minimum latency)
  • maxResults number: Maximum number of search results to return, 1-20. Default: 5
  • topic string: Category of the search. Options: "general" (default), "news" (real-time updates), "finance"
  • timeRange string (nullable): Filter results by publish date. Options: "day", "week", "month", "year". Default: null (no filter)
  • includeAnswer boolean: Include an LLM-generated answer to the query. Default: false
  • includeDomains array of strings: Domains to specifically include in results. Default: []
  • excludeDomains array of strings: Domains to specifically exclude from results. Default: []

Returns:

  • query string: The search query that was executed
  • answer string (nullable): AI-generated answer (only if includeAnswer is true)
  • results array: List of search results ranked by relevancy
    • title string: Title of the search result
    • url string: URL of the search result
    • content string: Content snippet from the search result
    • score number: Relevance score
    • rawContent string (nullable): Full parsed content (if requested)
  • responseTime number: Time in seconds to complete the request

Example Usage:

{
  "query": "latest AI funding rounds 2025",
  "searchDepth": "advanced",
  "maxResults": 10,
  "topic": "news",
  "timeRange": "week",
  "includeAnswer": true
}

Extract

Extract and parse web page content from one or more URLs. Returns cleaned content in markdown or text format, optimized for LLM consumption.

Operation Type: Query (Read)

Parameters:

  • urls array of strings (required): URLs to extract content from (max 20)
  • query string (nullable): Optional query for reranking extracted content chunks by relevance. Default: null
  • extractDepth string: Depth of extraction. "advanced" retrieves more data including tables and embedded content. Options: "basic" (default), "advanced"
  • format string: Format of extracted content. Options: "markdown" (default), "text"

Returns:

  • results array: Successfully extracted content
    • url string: The URL the content was extracted from
    • rawContent string: The extracted and cleaned page content
  • failedResults array: URLs that could not be processed
    • url string: The URL that failed
    • error string: Error description
  • responseTime number: Time in seconds to complete the request

Example Usage:

{
  "urls": [
    "https://techcrunch.com/2025/01/15/ai-startup-raises-100m",
    "https://en.wikipedia.org/wiki/Artificial_intelligence"
  ],
  "query": "funding amount and company details",
  "extractDepth": "basic",
  "format": "markdown"
}

Common Use Cases

Research and Analysis:

  • Search for recent news, academic papers, and industry reports to gather comprehensive research on any topic
  • Extract full article content from search results for deeper analysis and summarization

Competitive Intelligence:

  • Monitor competitor announcements, product launches, and pricing changes across the web
  • Extract structured data from competitor websites for comparison and benchmarking

Content Enrichment:

  • Enrich CRM records or lead data with real-time web information about companies and contacts
  • Search for and extract relevant content to support content creation and marketing workflows

Real-Time Monitoring:

  • Track breaking news and developments in specific industries or about specific companies
  • Monitor regulatory changes, policy updates, and market movements with time-filtered searches