io.github.purahmanian/google-trends-mcp

google-trends-mcp

Free Google Trends data inside any MCP-compatible AI client. No API key required.

UNOFFICIAL ENDPOINT DISCLAIMER

This server communicates with the same internal endpoints that the trends.google.com frontend uses. These endpoints are undocumented and unofficial. Google can change, rate-limit, or remove them at any time without notice. This project is not affiliated with, endorsed by, or sponsored by Google LLC. "Google Trends" is a trademark of Google LLC.

If requests start failing, open an issue. The fix is usually a URL or parameter tweak.

What it does

Quick start

No API key is needed. Install and run via npx.

Claude Desktop

Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "google-trends": {
      "command": "npx",
      "args": ["-y", "google-trends-mcp"]
    }
  }
}

Claude Code

claude mcp add google-trends -- npx -y google-trends-mcp

OpenAI Codex CLI

Add to ~/.codex/config.toml:

[mcp_servers.google-trends]
command = "npx"
args = ["-y", "google-trends-mcp"]

Example conversations

Spotting seasonality:

"Use Google Trends to show me interest in 'pumpkin spice' over the past 5 years worldwide."

The model calls interest_over_time with terms=["pumpkin spice"], timeframe="today 5-y" and narrates the clear September-October spikes each year.

Keyword competition:

"Compare search interest in 'Notion', 'Obsidian', and 'Roam Research' in the US this year."

The model calls compare_terms, reads the averages, and identifies the winner with supporting time-series data.

Opportunity discovery:

"What are the fastest-rising related queries for 'sourdough bread' right now?"

The model calls related_queries and surfaces the rising queries, often revealing adjacent niches and trending subtopics.

Development

# Clone and install
git clone https://github.com/purahmanian/google-trends-mcp.git
cd google-trends-mcp
npm install

# Run tests (no network calls; all HTTP is mocked)
npm test

# Build TypeScript to dist/
npm run build

# Run the server locally (stdio mode)
node dist/index.js

Rate limits and blocking

Google does not publish a rate limit for these endpoints. In practice, a few requests per minute works reliably. If you hit HTTP 429 or get blocked, the server returns a descriptive error message. Waiting 2-5 minutes usually resolves it.

Built by

Built by Puya Ventures LLC. I build custom MCP servers and AI integrations for product teams and researchers. Get in touch: purahmanian@gmail.com | Portfolio: puyarahmanian.com

Part of the Product-Research MCP Suite: keepa-mcp · google-trends-mcp · junglescout-mcp

Privacy

This server runs entirely on your machine. It collects no telemetry and stores no data. The only network calls it makes are to Google Trends endpoints (trends.google.com), sending the search terms and region codes you ask about. No account or API key is involved. See Google's privacy policy: https://policies.google.com/privacy

License

MIT. See LICENSE.