Brave Search MCP Server connects AI clients to the Brave Search API through the Model Context Protocol. It offers tools for web, local business, place, image, video and news search, LLM context for grounding and RAG, and AI summaries of search results.
It runs on your own machine with npx or Docker, over STDIO (the default) or HTTP. Setup guides cover Claude Desktop and VS Code. It needs Node.js 22 or newer to build from source, and a Brave Search API key. The code is open source under the MIT License.
The server is free; the Brave Search API is paid per use. Search plan: $5 per 1,000 requests. Answers plan: $4 per 1,000 requests plus $5 per million input/output tokens. Both include $5 in free credits every month. Enterprise plans are custom.
Developers who build chatbots, AI agents and coding assistants that need real-time web search results, and people who use MCP clients such as Claude Desktop or VS Code.
Has search tools for local businesses, places, images, videos and news, and an AI summary of search results; Firecrawl's tools centre on scraping, crawling and parsing pages.
Searches the whole web, news, images, videos and local businesses, not only library documentation.
Official Firecrawl MCP server: AI agents search, scrape and crawl the web
Scrapes pages into markdown or JSON, crawls and maps whole sites and parses PDFs; it has a hosted server, and a key-less mode for search, scrape and parse.
討論
登入以參與討論