MCP servers that ship as code — a repository or a package you install and run locally (usually over stdio). They have no hosted endpoint, so we cannot probe or grade them: instead each one shows its repository, install command, license, stars and when it was last updated.
Collected from GitHub repositories and npm, PyPI, Cargo, Go, Maven, NuGet and RubyGems packages tagged as MCP servers, the official MCP Registry, Docker's MCP catalog, ModelScope, Baidu MCP World, Cline and other public directories; one entry per repository. A server that also offers a hosted endpoint is listed under Remote instead. These entries are not counted in our reachability, verification or readiness statistics.
-
—
AGPL-3.0 -
tokenhub-websearchupdated this year
TokenHub WebSearch MCP Server - 使用 TokenHub 额度调用网络搜索
npm MIT -
websearch-mcpgithub.com/gawirable/websearch-mcpupdated this year
A self-hosted MCP server that gives AI agents deep internet research capabilities — no API keys required
-
mcp-websearch-servergithub.com/liliang-cn/mcp-websearch-serverupdated this year
A Model Context Protocol (MCP) server that provides multi-engine web search capabilities with content extraction.
★ 1 -
DeepSeek 官方 Search/Fetch API 的 MCP server(官方 SDK 构建,webSearch/webFetch 双工具,支持时效过滤与结构化输出)
npm MIT -
Cross-platform npm distribution of websearch-mcp-cli, a zero-API-key MCP search server (Go stdio binary) with per-platform prebuilt binaries.
npm MIT -
Multi-engine web search MCP server, CLI, and daemon — no API keys required. 16+ engines (Bing, Baidu, DDG, Brave, GitHub, arXiv, etc.), MDN / RFC / DevDocs / Context7 documentation, outline & JSON-LD structured article fetching.
Apache-2.0 ★ 3 -
A local DuckDuckGo web search MCP server. Exposes two tools via the MCP stdio protocol. Attempts to self rate-limit.
Apache-2.0 ★ 2 -
websearch-mcpgithub.com/sabari245/websearch-mcpupdated this year
MCP server for web search, fetch, and AI-powered extraction — powered by crawl4ai and morphllm
-
MyWebSearch - multi-engine web search MCP server, CLI and local daemon. No API keys required.
npm Apache-2.0 -
MCP web search server — DuckDuckGo Lite search via Model Context Protocol. Rust + rmcp.
MIT -
websearch-mcpgithub.com/iharee/websearch-mcpupdated this year
An agent-ready web information acquisition system that integrates multi-engine search and adaptive content fetching, exposed via MCP and LLM-friendly CLI.
MIT ★ 1 -
Local-first, zero-key web search MCP server: ddgs multi-engine + semantic search, reader-mode fetch, PDF, sqlite cache, GitHub search
MIT ★ 1 -
palena-websearch-mcpgithub.com/PalenaAI/palena-websearch-mcpupdated this year
🔍 Self-hosted web search MCP server for compliance-sensitive AI pipelines. Tiered scraping, PII detection (Presidio), pluggable reranking, content provenance, and full OpenTelemetry tracing. Built in Go for regulated industries.
Apache-2.0 ★ 1 -
Multi-engine web search + fetch MCP server (DDG, Brave, Tavily, Google AI Mode) with consensus-merge dedup, local cross-encoder reranking, and Groq synthesis
MIT ★ 1 -
多引擎联网搜索 MCP Server:国内引擎直连 + 免费节点池自运维 + 权威源优先;代理挂了自动降级直连。零 API Key、9 个工具、自带管理面板、466 个测试。Multi-engine web search MCP server, no API key required.
MIT -
MCP server that lets an agent search and scrape the web and pay for it itself — per request in USDC on Base via x402. No API keys.
npm MIT -
raw-web-searchgithub.com/andresfortunato/raw-web-searchupdated this year
Open source MCP search server for Claude Code. Replaces WebSearch + WebFetch.
MIT ★ 1 -
AgentWebSearch-MCPgithub.com/insung8150/AgentWebSearch-MCPupdated this year
Local WebSearch without API keys - MCP Server for Claude Code, OpenClaw, LM Studio. Parallel search across Naver/Google/Brave using Chrome CDP.
MIT ★ 22 -
web-retrieval-mcpgithub.com/velvetsp/web-retrieval-mcpupdated this year
MCP server for AI agents: high-fidelity web search (Exa) + tiered web fetch (Exa → optional local browser → Firecrawl) with an SSRF guard. A drop-in replacement for built-in WebSearch/WebFetch.
pypi MIT -
MCP server for AI agents: high-fidelity web search (Exa) + tiered web fetch (Exa→browser→Firecrawl) with SSRF guard. Drop-in replacement for built-in WebSearch/WebFetch in Claude Code, Codex, Grok, Cursor and all other agents.
MIT ★ 9 -
@mseep/searxngmcpstale
MCP server for performing internet searches using SearXNG with basic authentication
npm MIT -
A free (no API key needed) MCP server providing web search capabilities through SearXNG, optimized for LLM agents.
★ 6 -
Give you agents web search with Europe's own search provider, Staan, independent of Google, Bing and other American, Russian or Chinese search indexes.
MIT -
Fault-tolerant MCP Server for Stealth Web Search and Fetching
MIT ★ 4 -
Local Browser MCP
MIT ★ 1 -
duckduckgo-mcpgithub.com/T1ckbase/duckduckgo-mcpupdated this year
A MCP server for DuckDuckGo HTML search.
ISC -
Exa MCP for web search and web crawling!
MIT ★ 5,234 -
MCP server for enabling LLM applications to perform deep research via the MCP protocol
MIT ★ 379 -
An MCP server providing real-time web search capabilities to any AI model.
MIT ★ 16 -
MCP Server for Dumpling AI providing various Data scraping, conversion, and extraction tools
npm MIT -
@pseem/tavily-mcpstale
MCP server for advanced web search using Tavily
npm MIT -
@mseep/tavily-mcpstale
MCP server for advanced web search using Tavily
npm MIT -
SammelsuriumMCPgithub.com/winklermw-mz/SammelsuriumMCPupdated this year
This is a simple server application built using FastMCP, designed to handle various tools for retrieving information such as weather forecasts, web searches, Wikipedia queries, and more.
★ 1 -
CoexistAIgithub.com/SPThole/CoexistAIupdated this year
CoexistAI is a modular, developer-friendly research assistant framework . It enables you to build, search, summarize, and automate research workflows using LLMs, web search, Reddit, YouTube, and mapping tools—all with simple MCP tool calls or API calls or Python functions.
NOASSERTION ★ 531 -
RAG-Sandbox is a comprehensive, multi-tenant Agentic RAG (Retrieval-Augmented Generation) platform that redefines how users interact with their data. It moves beyond simple text retrieval to offer a fully multimodal, intelligent experience—transforming static documents into conversable agents, podcasts, and immersive stories.
MIT -
CoexistAIgithub.com/systemforge05/CoexistAIupdated this year
CoexistAI is a modular, developer-friendly research assistant framework . It enables you to build, search, summarize, and automate research workflows using LLMs, web search, Reddit, YouTube, and mapping tools—all with simple MCP tool calls or API calls or Python functions.
NOASSERTION -
A Model Context Protocol server with Exa for web search, academic paper search, and Twitter/X.com search. Provides real-time web searches with configurable tool selection, allowing users to enable or disable specific search capabilities. Supports customiz
npm mit -
A powerful native SillyTavern extension for high-performance local web search and scraping. Built using Playwright for superior page extraction with full MCP tool-calling parity. No external API keys required—fast, private, and fully integrated with the SillyTavern UI.
MIT ★ 5