Google Search trend data your assistant can call directly

Search demand is the default signal for briefs, SEO, and product questions. With source set to google search, an MCP-connected assistant returns dated points, normalized scores, and volume fields where the pipeline supplies them, so answers cite a time series instead of a static guess.

Why assistants default to Google Search first

Most user questions about demand, timing, or competition still anchor on web search. Feeding the model a scraped screenshot or a stale export trains hallucination risk. Trends MCP returns JSON the assistant can quote, chart, or diff. The contract matches the Data Sources table: pass source: "google search" and a keyword string.

Prompt patterns that route to the tool

Docs suggest including "using TrendsMCP" or "via TrendsMCP" so the client selects the MCP tool instead of a web search. Example: "Via TrendsMCP, show Google Search growth for GLP-1 over 12 months." That mirrors the AI prompt examples shipped beside the schema.

Pairing history with live discovery

Use get_trends when the user needs shape and seasonality. Use get_growth when they need a decision score across windows. Use get_top_trends with type Google Trends for intraday culture questions. Ranked lists from get_ranked_trends help when the brief says "what queries are accelerating" without naming a keyword.

Related product pages

Deeper positioning for search-specific buyers lives on Google Search trends and Google Search data. The cross-tool overview sits on MCP trend tools for assistants.

Setup reminder

Add the hosted MCP URL and bearer key from trendsmcp.ai, then keep the API reference open for quota rules and error codes.

get_trends

Load multi-year weekly demand before the assistant writes a memo or content outline.

get_trends(keyword='heat pump', source='google search', data_mode='weekly')

get_growth

Return 3M, 12M, or YTD percent change for prioritization without manual spreadsheet work.

get_growth(keyword='heat pump', source='google search', percent_growth=['3M', '12M', 'YTD'])

get_ranked_trends

Surface fast-rising queries on Google Search when the user asks what is moving in a category.

get_ranked_trends(source='google search', sort='wow_pct_change', limit=30)

get_top_trends

Answer what is trending on Google right now using the live Google Trends feed.

get_top_trends(type='Google Trends', limit=25)

Common questions

Use google search in lowercase with a space, as listed in the Data Sources table on trendsmcp.ai/docs. Typos trigger invalid_source errors, so copying from the docs beats paraphrasing.
Any keyword or phrase the docs allow for that source: brand names, product categories, long-tail questions, and competitor terms. The assistant should pass the same string a human would type into Google.
For supported Google-style modes, weekly pulls a long history window; daily focuses on recent days. Interpret returned dates as source-native cadence. When in doubt, start weekly for macro context then add daily for the last month.
Yes. get_growth accepts comma-separated sources or all so the model can compare google search with youtube, tiktok, or amazon in one response for the same keyword where formats align.
Get Google Search trend data your assistant can call directly in 30 seconds
Free tier includes 100 requests per month. No credit card required. Works with Claude, Cursor, ChatGPT, Raycast, and every MCP client.
Get your free API key