The Tavily Search Node is a powerful web-search engine connector that allows your AI agent to query the web in real-time, fetching highly relevant snippets, summaries, and source URLs. It is optimized to extract clean information without bloating the context window with raw HTML.
Parameters
Required Parameters
- query (string): The search query to look up on the web.
Optional Parameters
- country (string): Boost search results from a specific country (e.g., 'United States', 'Germany', 'Japan'). Must be full country name; ISO codes are not supported.
- search_depth (string): Depth of search. Available options:
basic: Standard search results.advanced: More thorough, deep search.fast: Optimized low latency with high relevance.ultra-fast: Latency-prioritized low latency search.
- max_results (integer): The maximum number of search results to return (Default:
5). - topic (string): Category of search. Choose
generalornews. - time_range (string): Returns results from a specific time window back from the current date.
- start_date / end_date (string): Filter results after/before a specific date in
YYYY-MM-DDformat. - exact_match (string): Only return results containing this exact phrase.
- include_domains / exclude_domains (array): Filter results to include or exclude specific site domains.
- include_favicon (boolean): Include favicon URLs for each source.
- include_images (boolean): Include query-related images in the response.
- include_image_descriptions (boolean): Include text descriptions for returned images.
- include_raw_content (boolean): Include cleaned and parsed full HTML content for each result.
Practical Use Cases
- Real-Time Fact Checking: Automatically retrieve current information or news to cross-verify facts during conversational flows.
- Competitive Intelligence: Monitor brand mentions or look up latest competitor announcements across specific included domains.
- News Summarization: Perform hourly checks on specific topics with
topic: "news"andtime_rangeoptions to feed a newsletter generator.