Start free

Platforms MCP server

MCP server · 3 tools · 1 endpoint

Google News MCP server

Keyword search across news headlines, at 1 credit an article, so an agent knows what happened this morning.

Get a free key Connect in a minute

100 credits on signup · No card · Sign in with OAuth or paste a key

scrapersocial-google_news mcp.scrapersocial.com/google_newsconnected
you

Search Google News for "<topic>" and tell me whether coverage is mostly positive or negative.

agent
calls search 1 credit / article
{ "query": "<keywords>" }
server
returns
articles [{ … }] · one per article
Shape only. Keys follow the tool's summary; a real call returns the full record. 1 call · 1 credit

Connect

One URL. Your client already knows what to do with it.

The first call opens a sign-in. New accounts get 100 credits, no card. Every client below shares one account and one balance.

Get a free key
server url https://mcp.scrapersocial.com/google_news
claude.aiOAuth

Settings → Connectors → Add custom connector, then paste the server URL.

mcp.scrapersocial.com/google_news

Web, desktop and mobile. Nothing else to paste.

Claude Codeone command

Run this in a terminal.

$ claude mcp add --transport http scrapersocial-google_news https://mcp.scrapersocial.com/google_news

Approve the sign-in prompt on first use.

Cursorone click

Installs the server and opens the MCP panel.

Add to Cursor
or paste the JSON into .cursor/mcp.json
{ "mcpServers": { "scrapersocial-google_news": { "url": "https://mcp.scrapersocial.com/google_news", "headers": { "Authorization": "Bearer sk_live_..." } } } }
Other clients · VS Code, Windsurf, Gemini CLI, anything that speaks MCP
VS Codeone click or JSON
Add to VS Codeor run MCP: Add Server, choose HTTP and use:
{ "name": "scrapersocial-google_news", "type": "http", "url": "https://mcp.scrapersocial.com/google_news" }
Windsurf~/.codeium/windsurf/mcp_config.json
{ "mcpServers": { "scrapersocial-google_news": { "serverUrl": "https://mcp.scrapersocial.com/google_news", "headers": { "Authorization": "Bearer sk_live_..." } } } }
Gemini CLI~/.gemini/settings.json
{ "mcpServers": { "scrapersocial-google_news": { "httpUrl": "https://mcp.scrapersocial.com/google_news", "headers": { "Authorization": "Bearer sk_live_..." } } } }
Any clientStreamable HTTP · OAuth 2.1 or Bearer
$ npx add-mcp https://mcp.scrapersocial.com/google_news

One account and one balance across every client and the every-platform hub at mcp.scrapersocial.com.

What it does

What your agent can do with Google News

One endpoint, one job. Behind https://mcp.scrapersocial.com/google_news there is keyword search across news headlines, returning the results as a list your agent can read: the headline, the outlet and the link, 1 credit per article. That is the entire surface, and the simplicity is the point. A model with a training cutoff can’t know about this week’s announcement; give it this tool and it can check.

The capability tool is search. Pass a query, get articles back. list_endpoints is free and will confirm there is a single path, /v1/google_news/search, and call_endpoint runs it with the same query parameter if your client prefers calling by path. There is nothing else to learn.

Google News is an aggregator, not a social platform. ScraperSocial includes it alongside Reddit, X and the rest because the question “what is being said about ” has a press component and a social component, and an agent that can only answer half of it gives a lopsided answer. Headlines arrive in the same envelope as everything else on the hub, so the model can read coverage and conversation in one pass without you writing a second integration.

Four jobs it finishes on its own

01

Current-events grounding before answering

A user asks about a company, a bill or a person in the news. The agent runs search first, reads the headlines, then answers with dates and sources instead of stale recall. Ten articles is 10 credits.

02

Brand coverage monitoring

Query your brand and your competitors' names on a schedule and have the model note new outlets, new angles and anything negative. Deduplication across runs is yours to build; the search itself is cheap.

03

Research briefs with a press section

When compiling a brief, pull the latest headlines for the topic and let the model summarise the coverage, with links the reader can verify. One search of 50 articles costs 50 credits and gives plenty to work from.

04

Crisis awareness for support teams

An agent handling support can check whether an outage, recall or incident involving the company is in the news before it replies. A single query per conversation keeps the cost negligible.

Limits · read before you connect

What this does not do

Written by a person, not generated. If something here is wrong, tell us.

Headlines, sources and links only; the article body is not returned, so fetch the page yourself if you need the text. Search is a point-in-time query with no persistent alert, no deduplication and no notification: scheduling is your side of the line. A result list holds 20 articles until you raise the limit, and the limit tops out at 50; a next_cursor in meta picks up where the last page stopped. There is no filter by outlet, language or date in the catalogue; narrow with the query itself. Nothing is written anywhere and no Google account is involved. A query with no matching articles is still charged 1 credit.

Tools

3 tools, same names as the every-platform hub

One does one Google News thing each. The last two reach all 1 endpoint by path, so nothing in the catalog is out of an agent's reach.

Full credit table →

Search a platform

search

Keyword search over one platform's public content.

1 credit / article
Reaches every endpoint by path

List every Google News endpoint

list_endpoints

Browse every Google News endpoint, its input and its credit cost. Free, and calls nothing upstream.

free

Call any Google News endpoint by path

call_endpoint

Run any Google News endpoint by its path, so nothing Google News offers is out of reach of an agent.

endpoint's own cost

Pricing

100 free credits is 100 search results.

Then $5 a month for 1,000 credits, or $4 per 1,000 as a top-up. No card to start. Every call costs 1 credit.

100 search resultssearch · 1 each
Get a free key 100 credits on signup · $5 for 1,000 · full credit table

Prompts

Things to ask once it is connected

Copy one, swap the placeholder, paste it into your client.

What is the latest news about <company>? Summarise the top ten headlines with sources and links.

Search Google News for "<topic>" and tell me whether coverage is mostly positive or negative.

Has <product> been in the news this week? List any recalls, outages or incidents.

Find recent headlines mentioning <person> and group them by outlet.

Give me a press roundup on <industry term> I can paste into a Monday briefing.

Questions

Before you connect

What does a news search cost? #

1 credit per article returned. The default page is 20 articles, the maximum is 50. The free 100 credits cover five full pages, and $5 a month buys 1,000 credits after that.

Does it return the full article text? #

No. You get the headline, the source and the link. If the agent needs the body, it has to fetch the URL separately; that is outside this server.

Can I filter by date, language or outlet? #

Not with a parameter. The single endpoint takes a query and a limit. Put the outlet or a date term in the query if you need to narrow, and let the model drop what it doesn't need.

How is this different from the Google Search spoke? #

Different index and different unit. This server searches the news index and returns articles; the Google Search server returns web result pages and Business Profile data. They're separate spokes with separate paths.

Is ScraperSocial part of Google? #

No. ScraperSocial has no affiliation with Google; it's an independent company that reads public headlines and nothing more.