Serply
Google web, news and Scholar search from any AI agent via the Serply API. 3 tools, API-key auth.
7 天免费 · 无需信用卡 · EU 托管
问 Serply 任何事
通过 AnythingMCP 对 Serply 试一个真实提示 — Claude 调用工具并返回答案。在你的消息上按 Enter 即可在 Cloud 安装。
在 Cloud 上安装 SerplyClaude 是 AI,可能会出错。请核对回答。
2 分钟设置
- 1在 cloud.anythingmcp.com 开始免费试用。
- 2在连接器商店点击 Serply。提示时粘贴凭据。
- 3生成 MCP API 密钥,将 AI 客户端指向生成的 MCP URL。
Serply 指南
面向主流 AI 客户端的逐步设置。
配置 Serply
直接来自连接器定义:在哪里获取凭据、需要哪些权限,以及该适配器调用的是哪个 API。 由连接器作者以英文提供。
This connector wraps the Serply REST API (api.serply.io), which returns Google search results as JSON.
Setup:
- Create an account at https://serply.io (2,500 free credits, no card required) and copy the API key from the dashboard.
- Set
SERPLY_API_KEY.
Authentication: the key is sent as the X-Api-Key header on every call, so it never appears in a logged URL.
Three tools, two endpoints:
serply_search- Google web search,GET /v1/search. Returns organic hits underresults[](title,link,description), plusanswers,knowledge_graphandrelated_questionswhen Google shows them.serply_news- the same endpoint with Google's news vertical (tbm=nws), so news hits come back in the sameresults[]shape with direct publisher links.serply_scholar- Google Scholar,GET /v1/scholar. Returns papers underarticles[](title,link,author,description).
Parameters:
qaccepts Google operators:site:,-site:,filetype:,inurl:,intitle:"...", quoted phrases.numis an upper bound, capped at 10 per request by Serply. At 10 the row count follows the Google page and can be 9 or 11.glis a two-letter country code (us,gb,de). Leave it unset and the region is whatever Google assigns the request, so set it when locality matters.tbsrestricts recency on web and news:qdr:h(hour),qdr:d(day),qdr:w(week),qdr:m(month),qdr:y(year). Custom date ranges are not supported.
Result shape: each hit carries a URL, title and snippet. Snippets are previews; for exact values on a page, fetch the result link separately. This connector does not extract page content.
Pricing: one credit per request. Docs: https://serply.io/docs