Source: Tavily API docs (Search endpoint and credits)

Tavily's reference for its LLM-oriented search API: search depths, chunked results, raw content, domain/time filters and credit pricing for search, extract, map, crawl and research.

Contents
  • What: official API documentation (two pages)
  • Author / publisher: Tavily
  • Date: read 2026-11-14 (living pages)
  • Original: Search endpoint, API credits

Summary

Tavily is a search API designed to feed LLMs: one call returns ranked results with pre-extracted, query-relevant content chunks, optionally full page text and an LLM-written answer.

Key claims

  • POST https://api.tavily.com/search with a bearer key.
  • search_depth: ultra-fast, fast, basic (default), advanced; advanced costs 2 credits, the rest 1.
  • chunks_per_source 1-3 (default 3); max_results 0-20 (default 10); topic general/news/finance; time_range day/week/month/year; include_domains up to 300; include_answer (basic/advanced); include_raw_content (markdown/text).
  • auto_parameters lets Tavily choose settings from query intent and may upgrade to advanced (2 credits).
  • Each result has a relevance score, a content snippet built from chunks, plus optional raw_content, published_date, images.
  • Other endpoints: Extract (1 credit per 5 URLs basic, 2 advanced), Map (1 credit per 10 pages), Crawl (map + extract), Research (mini 4-110 credits, pro 15-250 credits per request).
  • Pricing: 1,000 free credits per month; pay-as-you-go $0.008 per credit; plans $0.0075-$0.005 per credit. Failed extractions are not charged.

Feeds

Web search APIs

Suggest changes →