VX Agents

OpenAPI Marketplace · Search

Brave Search API

v1.0.1 · 6 endpoints · 1 connections · shared by linh

Official OpenAPI specification for the Brave Search API. Provides independent, privacy-preserving web search, LLM context grounding for AI agents and RAG pipelines, news, images, videos, and query suggestions. **Features:** - Privacy-first web index independent of Google and Microsoft. - Specialized LLM Context grounding endpoint (`/llm/context`) designed for AI agents. - Rich multimedia search across news, images, and videos. - Custom ranking via Brave Goggles and freshness filters.

Connect this API to an AI agent Browse all specs

What you can do with Brave 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.

MethodPathOperationDescription
GET /web/search searchWeb Search the Web
GET /llm/context getLlmContext LLM Context Grounding
GET /news/search searchNews Search News Articles
GET /images/search searchImages Search Images
GET /videos/search searchVideos Search Videos
GET /suggest/search suggestQueries Query Autosuggest

Authentication

ApiKeyAuth

API key — sent as the X-Subscription-Token header

Brave Search API Subscription Token. Sign up and generate a token at https://brave.com/search/api/. Pass this token in the X-Subscription-Token header on every request.

Usage guide

Brave Search API Usage Guide for AI Agents

Authentication

Every request requires your Brave Search API key sent in the header:

X-Subscription-Token: <YOUR_BRAVE_SEARCH_API_KEY>

Get an API key at https://brave.com/search/api/.

Versioning

To pin an exact API release, send the optional header:

Api-Version: YYYY-MM-DD

1. Web Search (/web/search)

Use for general search queries to retrieve organic web results, snippets, discussions, and infoboxes.

GET /res/v1/web/search?q=artificial+intelligence&count=5&freshness=pw

Key Parameters:

  • q: Query string (max 600 characters).
  • count: Number of results (1-20, default: 20).
  • freshness: Filter by recency: pd (past 24h), pw (past week), pm (past month), py (past year).
  • extra_snippets: Set to true to get up to 5 additional text excerpts per result for deeper context.
  • units: metric or imperial.
  • goggles: Custom Goggle URL or ID for community-driven reranking.

2. LLM Context Grounding (/llm/context)

Optimized specifically for LLM prompts and Retrieval-Augmented Generation (RAG). Returns concise, structured text snippets with high signal-to-noise ratio.

GET /res/v1/llm/context?q=latest+developments+in+quantum+computing

3. Specialized Search

  • News Search (/news/search): Real-time news articles with publish timestamps.
  • Image Search (/images/search): Web images with thumbnails, dimensions, and source pages.
  • Video Search (/videos/search): Video results with creators, views, and durations.
  • Autosuggest (/suggest/search): Query completions for interactive agent search bars.

How to use this spec

  1. Create a free VX Agents account (sandbox tools included).
  2. Open the marketplace and connect “Brave Search API” to an agent.
  3. Expose it as an MCP server for Claude/ChatGPT, or chat with it directly.

Related APIs