Tavily Search API
Official OpenAPI specification for the Tavily Search API (`api.tavily.com`). Tavily is an AI-first search and web intelligence engine built specifically for autonomous AI agents, LLM pipelines, and RAG systems. **Key Features:** - AI-curated search results with extracted relevance snippets, reducing LLM context window bloat. - Built-in LLM answer synthesis (`include_answer: true`) for rapid direct answers. - Comprehensive URL extraction (`/extract`), website structure mapping (`/map`), and deep site crawling (`/crawl`). - Domain filtering (`include_domains`, `exclude_domains`), news/finance topics, and time freshness filtering.
Connect this API to an AI agent Browse all specs
What you can do with Tavily Search API via MCP
Every endpoint below becomes a governed tool your AI agent (Claude, ChatGPT, or a customer-facing VX agent) can call — with scoped permissions, sandbox testing, and audit logs.
| Method | Path | Operation | Description |
|---|---|---|---|
| POST | /search |
searchWeb | Search the Web |
| POST | /extract |
extractContent | Extract Content from URLs |
| POST | /map |
mapWebsite | Map Website Structure |
| POST | /crawl |
crawlWebsite | Crawl and Extract Website |
Authentication
ApiKeyAuth
Bearer token — sent as Authorization: Bearer <token>
Tavily API key passed as Bearer token in Authorization header (Authorization: Bearer tvly-...).
Alternatively, pass api_key in the request body.
Usage guide
Tavily Search API Usage Guide for AI Agents
Authentication
Authenticate by passing your Tavily API key as a Bearer token in the Authorization header:
Authorization: Bearer tvly-YOUR_API_KEY
(Alternatively, you may pass "api_key": "tvly-..." directly in the JSON request body).
Get a free API key (1,000 requests/month, no credit card required) at https://tavily.com.
1. AI Web Search (POST /search)
General web search with real-time AI snippets and optional answer generation.
POST https://api.tavily.com/search
Content-Type: application/json
Authorization: Bearer tvly-YOUR_API_KEY
{
"query": "What are the latest updates on quantum computing?",
"search_depth": "basic",
"topic": "general",
"max_results": 5,
"include_answer": true
}
2. Extract Webpage Content (POST /extract)
Extract clean markdown/text content from specific URLs without boilerplate or ads.
POST https://api.tavily.com/extract
Content-Type: application/json
Authorization: Bearer tvly-YOUR_API_KEY
{
"urls": ["https://en.wikipedia.org/wiki/Artificial_intelligence"],
"extract_depth": "basic",
"include_images": false
}
3. Site Mapping & Crawling (POST /map & POST /crawl)
- Map (
/map): Discovers all linked paths under a root URL. - Crawl (
/crawl): Discovers and extracts full page content across a target website.
How to use this spec
- Create a free VX Agents account (sandbox tools included).
- Open the marketplace and connect “Tavily Search API” to an agent.
- Expose it as an MCP server for Claude/ChatGPT, or chat with it directly.