Free Search MCP
未认领Local-first, no-API-key MCP server for web search, fetch, and document reading. Multi-engine (DuckDuckGo, Mojeek, Startpage), smart Playwright fallback, FTS5 cache, LLM-tuned output.
安装
claude mcp add search -- uvx free-search-mcp # Claude Code工具（29 个）
anysearch
AnySearch
bilibili
keyless Bilibili (哔哩哔哩) video search via the public
brave
, baidu — intermittent challenges to headless clients.
brave_api
Brave Search API
category
news / pdf / github / paper / forum / blog
dataset
zenodo
exclude_domains
["pinterest.com"]
exclude_text
"beginner"
forum
stackexchange, hackernews
freshness
day / week / month / year
github
github (repos + issues/PRs), githubcode (needs a token)
github_code
GitHub
keyless Google web SERP scrape (HTTP first, Playwright
google_cse
Google Custom Search
image
openverse
include_domains
["python.org", "djangoproject.com"]
include_text
"async"
max_age_hours
24
news
googlenews, gdelt
openlibrary
book search over the Internet Archive catalogue.
paper
arxiv, openalex, crossref, pubmed
searx
meta-search proxy via public SearXNG instances; included for
serper
Serper (Google)
sogou
, so360 — Chinese web indexes, HTML scrapes, best-effort like
stackexchange
Stack Apps (key optional)
startpage
browser-rendered (~5-10s/query); good for hard-to-reach
tavily
Tavily (AI search)
wikipedia
encyclopedia search; language follows SEARCHMCPREGION.
zhihu
best-effort keyless Zhihu (知乎) search. Zhihu's