AI Agents
brightdata-mcp
Bright Data's MCP server: web search, scraping, structured data and browser automation for agents (75 tools)
★ 2.7k starsMIT
View guide →Brave's official MCP server: web, news, image, video, local, and LLM-context search for your agents.
brave-search-mcp-server is an open-source ai agents project by brave (brave/brave-search-mcp-server on GitHub). It has 1,489 GitHub stars and 207 forks, is released under the MIT license, and was last updated on GitHub on 5 October 2026.
Decide which Brave Search plan and tools you need; local search's full capabilities require the Pro plan.
Sign up for a Brave Search API account and generate an API key from the developer dashboard.
Add the server to your MCP client using the npx or Docker configuration, passing your key as BRAVE_API_KEY.
Restart the client and ask the agent to search the web, find local places, or fetch LLM-ready context.
Web, local, place, image, video, and news search, plus a summariser and an LLM-context tool for grounding.
Country, language, result counts, pagination, safesearch levels, freshness ranges, spellcheck, and custom Goggles re-ranking on the main tools.
Returns pre-extracted page content with token, snippet, and URL budgets tuned for RAG pipelines and agent grounding.
STDIO by default with an HTTP mode (configurable host, port, and stateless operation), distributed via npx or Docker.
Environment variables and CLI flags control enabled or disabled tools, log level, allowed origins and hosts for HTTP, and key loading from a file or Docker secret.
Documented setups for Claude Desktop, VS Code, and other MCP clients, plus MCP Inspector testing workflows.
The npx route needs no clone; a local build uses npm.
git clone https://github.com/brave/brave-search-mcp-server.git
cd brave-search-mcp-server
npm install && npm run buildAdd the server to your MCP config with your key, using the STDIO transport as in the README's VS Code example.
# mcp.json: command "npx", args ["-y", "@brave/brave-search-mcp-server", "--transport", "stdio"], env BRAVE_API_KEY=<your key>After restarting, ask your agent a question that requires current web information. # see README usage sections
# verify with /mcp list in clients that support it, per READMEIt is maintained by Brave, the company behind the Brave browser and the Brave Search API.
The server code is MIT-licensed and free; the Brave Search API it calls has its own plans, including the search and answers tiers described in the README, and some features require the Pro plan.
Yes: a Brave Search API key is required and is passed as BRAVE_API_KEY (or loaded from a key file) for every configuration.
Tavily-style tools focus on extraction for AI; Brave's server is the first-party interface to Brave's independent search index, adding local and place search plus summarisation alongside web results.
Get a Brave Search API key, add the npx configuration to your MCP client with BRAVE_API_KEY set, restart the client, and search.
Bright Data's MCP server: web search, scraping, structured data and browser automation for agents (75 tools)
Official MathWorks MCP server: drive MATLAB from AI coding agents.
Exa's MCP server for web search and crawling — give your AI agent a live, high-quality web index.
Simple desktop AI assistant with knowledge bases, agents, and web search.
Qwen-AgentWorld: a language world model simulating MCP, terminal, web, and OS environments for agents.
A local MCP server for serious arXiv work — LaTeX section reads, BibTeX and topic watches.
The most-starred other AI Agents projects in the LibreAI catalogue — the realistic shortlist if brave-search-mcp-server is not the right fit.
A personal AI that really does things — on any OS, with your data staying yours.
An agentic skills framework & software development methodology that works.
Agent harness optimization system with skills, memory and security for Claude Code, Codex and more
The agent that grows with you — a Python framework for adaptive AI agents.
DeepSeek's plugin-based agent harness where everything is a plugin.
Free and open — grab the source, star the repo, and start creating.