google-data-mcp
MCP server providing access to Google Trends, YouTube listings, Google Ads Transparency, and Google Play reviews via plain HTTP, with no API key required. It caches results and paces requests to work within Google's informal rate limits.
README
google-data-mcp
An MCP server for four Google data surfaces that answer to plain HTTP — no API key, no browser, no scraping service:
| Tool | What it returns |
|---|---|
google_trends_interest |
Interest over time for one keyword (0–100, weekly or hourly points) |
google_trends_compare |
Up to 5 keywords in ONE request, so the values are comparable |
google_trends_trending |
What is trending right now in a country |
youtube_listing |
A channel's Shorts, a channel's videos, or search results |
google_ads_advertisers |
Advertisers in the Ads Transparency Center, by name |
google_ads_creatives |
Every ad an advertiser runs — format, preview URL, first/last shown |
google_play_reviews |
App reviews — rating, text, date, version, developer reply |
Read this before you install
Everything leaves from your machine's IP, and Google's per-IP budget accumulates over hours. This is not a caveat buried at the bottom; it is the main thing that decides whether this server is right for your job.
Measured: an unpaced burst was rate-limited at request 93, and after roughly 130 paced requests spread over hours the same address was refused on the first request of a fresh session. While writing this server, both the author's IPs — a home connection and a datacenter VPS — ended the day refused by Google Trends.
So this server caches every result on disk (~/.cache/google-data-mcp) and paces its requests. That
is enough for interactive use: asking an agent a handful of questions, exploring a topic,
checking a competitor. It is not enough for bulk work, and no amount of local code can make one
IP behave like many.
If you need volume, the same clients run behind rotating proxies as Apify Actors: Google Trends · YouTube · Ads Transparency.
What is deliberately missing
YouTube transcripts. They look available — the watch page still lists caption tracks — but
api/timedtext returns zero bytes and /youtubei/v1/get_transcript answers Precondition check failed even when sent the page's own INNERTUBE_CONTEXT and visitorData, and the ANDROID and IOS
player clients are refused the same way. A transcript tool here would be a promise this server
cannot keep, so there isn't one.
Install
pipx install google-data-mcp
# or: pip install google-data-mcp
# or, without installing anything: uvx google-data-mcp
Then register it with your MCP client. For Claude Code:
claude mcp add google-data -- google-data-mcp
Or by hand, in an MCP client config:
{
"mcpServers": {
"google-data": {
"command": "google-data-mcp"
}
}
}
Notes on the data
- Trends values are relative within a single request. Two separate
google_trends_interestcalls are not comparable to each other; that is whatgoogle_trends_compareis for. Google caps a comparison at 5 terms and silently drops a 6th, so a 6th is refused rather than quietly ignored. - Play reviews differ per locale. Each
hl/glpair returns a different set of reviews — six locales gave 120 unique reviews of the same app. Vary them to widen coverage rather than paging deeper in one language. - Reviewer identity is omitted by default. The records carry an author name, avatar and Google
account id; rating, text, date and version answer product questions without them. Set
include_authorif you genuinely need the name. - Zero ads is a real answer. Google keeps advertisers whose ads have stopped running and still
reports a count for them, so an empty
adslist with a non-zero declared count is information, not a failure. searchVolumeon trending terms is Google's own rounded bucket, not a precise count.
Legal
These are public pages, but none of them has an official public API and each platform's Terms of Service restrict automated access. You are responsible for how you use the output. No personal data is collected by default.
Development
python -m venv .venv && .venv/bin/pip install -e .
.venv/bin/python -m google_data_mcp # speaks MCP over stdio
Each client is plain standard library and can be exercised on its own, which is the fastest way to check whether Google changed something:
.venv/bin/python -c "from google_data_mcp.play import PlayClient; print(len(PlayClient().reviews('com.spotify.music', limit=40)))"
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
Neon Database
MCP server for interacting with Neon Management API and databases
E2B
Using MCP to run code via e2b.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.