Tavily Search API

Tavily Search API provides real time web search, ranked content snippets, domain and date filtering, localization, images, and structured results for AI agents and applications.

About Tavily Search API

Tavily Search API is a web search API designed for developers and AI applications. It searches the web, ranks relevant sources, extracts useful content, and returns structured results that can be used by AI agents, LLM applications, research systems, and other software requiring current web information.

The /search endpoint accepts POST requests with JSON input and supports multiple search depths, including basic, advanced, fast, and ultra-fast modes. Search requests can return up to 20 results and can be configured by topic, date range, included or excluded domains, country, language, safe search settings, exact matching, images, favicons, raw content, and generated answers.

Search results can contain a title, URL, extracted content, relevance score, raw content when requested, publication date when available, favicon, images, response time, request ID, and usage information. Tavily authenticates API requests with an API key supplied as a Bearer token in the Authorization header.

Key Features

Web Search Endpoint: The /search endpoint retrieves and ranks web results for a supplied query and returns structured search data.
Multiple Search Depths: Supports basic, advanced, fast, and ultra-fast search modes for different relevance and latency requirements.
Result Filtering: Supports date ranges, publication date filtering, included domains, excluded domains, country boosting, language controls, and safe search.
Content Extraction: Search results include relevant content snippets, with configurable chunks per source for supported search depths and optional raw page content.
Rich Search Responses: Requests can optionally include generated answers, images, image descriptions, favicons, publication dates, and usage information.
Developer Access: Supports direct REST requests as well as official Python and JavaScript SDK usage.

Pricing

Researcher: Free/month Includes 1,000 API credits per month, no credit card requirement, and email support.
Pay As You Go: $0.008/credit Usage based option where customers pay for the API credits they consume.
Enterprise: Contact the sales team for pricing information. Includes custom API calls, custom rate limits, enterprise support and SLAs, security and privacy features, and custom seats.

Search requests consume credits according to search depth. Basic, fast, and ultra-fast searches consume 1 API credit per request, while advanced searches consume 2 API credits per request.

Use Cases

Add current web search to AI agents and LLM applications
Retrieve ranked web context for retrieval and grounding workflows
Build research and information retrieval applications
Search selected domains or time periods for relevant web content

Pros & Cons

Pros:

Supports up to 20 ranked web results in a single search request
Provides extensive controls for domains, dates, language, country, topic, content, and response data
Offers REST access plus official Python and JavaScript SDKs

Cons:

A search request can return a maximum of 20 results
Country based result boosting is available only when the search topic is set to general

Integrations

REST API, Python SDK, JavaScript SDK, cURL

FAQ

What does Tavily Search API do?

Tavily Search API searches the web, ranks relevant sources, extracts useful content, and returns structured results for developers and AI applications.

How is Tavily Search API authenticated?

Requests are authenticated using a Tavily API key supplied as a Bearer token in the Authorization header.

How many search results can Tavily return?

The max_results parameter supports up to 20 results in a search request.

What search filtering options does Tavily support?

Tavily Search supports date ranges, publication date filtering, included and excluded domains, country boosting, language controls, topics, exact matching, and safe search.