Source: Gemini API Grounding with Google Search docs
Google's guide to the google_search tool in the Gemini API: model-driven queries, url_citation spans, search suggestions and per-query pricing.
Contents
- What: official API documentation
- Author / publisher: Google
- Date: read 2026-11-14 (living page)
- Original: aistudio.google.com/docs/google-search (formerly ai.google.dev/gemini-api/docs/google-search)
Summary
Enabling the google_search tool lets Gemini decide whether to search, generate its own queries, read results and return an answer with inline citations.
Key claims
- The response contains a
google_search_callstep (queries run), agoogle_search_resultstep withsearch_suggestions(HTML to render in the UI, governed by the terms of service), and model text withurl_citationannotations carryingurl,title,start_index,end_index. - Inline citations are built by slicing the text with each annotation’s indices.
- Pricing (Gemini 3 and newer): billed per search query executed; 5,000 free per month, then $14 per 1,000 queries; empty queries not billed. Gemini 2.5 and older: billed per prompt. Default paid-tier limit 1,500 requests per day per project.
- Supported across Gemini 2.0 to Gemini 4 models; older models use
google_search_retrieval. - Combines with code execution, URL context and Google Maps grounding; Gemini 3+ can mix built-in tools with custom function calling.