Google Trends MCP servers compared, with a real analysis in Claude

The main ways to get Google Trends data into Claude, ChatGPT or Cursor through MCP are Cite42, HasData's Google Trends MCP, SerpApi's MCP server, DataForSEO's MCP server, Trends MCP and Apify's Google Trends Scraper. We found no Trends MCP server published by Google as of October 2026. Cite42 is a hosted server at https://mcp.cite42.dev with OAuth sign-in, and each Trends request costs $0.05, billed per call with no subscription.

The second half of this guide uses Cite42 in Claude to compare a year of Google Trends data for the US and UK, including the trend label an assistant is most likely to misread.

Cite42 publishes this guide. The worked example uses two real Cite42 requests from October 5, 2026. We researched the other servers from their repositories, documentation and pricing pages on October 6, 2026, without installing them.

Google Trends MCP servers compared

• Cite42: one term per request with interest over time, related queries and a trend label, plus keyword volume and AI visibility tools in the same server. $0.05 per request, no subscription.
• HasData Google Trends MCP: the fullest single Trends tool, with interest over time, interest by region down to city level, related queries with growth percentages, and related topics.
• SerpApi MCP: one search tool across every SerpApi engine, including Google Trends with up to five terms per chart.
• DataForSEO MCP: a general tool that reads DataForSEO's docs and calls its API endpoints, including Google Trends Explore, useful if you already have a DataForSEO account.
• Trends MCP: Google Search trends alongside YouTube, TikTok, Reddit, Amazon and other sources (35+ in all), with growth percentages.
• Apify Google Trends Scraper: a scraper exposed through Apify's hosted MCP server, billed per result.

Pick HasData or SerpApi when you need region maps or related topics, SerpApi when you want several terms on one chart, and Trends MCP when you want Google next to TikTok or YouTube. MCP directories such as Glama also list free community servers that call Google's unofficial endpoints directly. Those can break or get rate limited the same way pytrends did, so check which data source a server uses before you rely on it.

Cite42: connect the hosted server

Cite42 is a pay-per-call API and MCP server for Google Trends, keyword search volume and AI visibility. Its cite42_trends tool returns one term's interest series, a rising, falling or stable label and related queries for $0.05 per request. There's no subscription or monthly fee: you top up a prepaid balance from $25.00, the balance never expires, you pay only for successful calls, and a verified signup includes $1.00 of free credit.

In Claude Code, add the server, then sign in with claude mcp login cite42:

claude mcp add --transport http --scope user cite42 https://mcp.cite42.dev

In the Claude app (as a custom connector), ChatGPT (developer mode), Codex or Cursor, add https://mcp.cite42.dev as a remote MCP server and sign in when prompted. The setup docs cover each client, and the Claude Code SEO guide walks through the full Claude Code and Codex setup.

The trade-off is scope: one term per request, with no multi-term chart or region map, and values on Google's relative 0 to 100 index.

Ask for one term, one market and one window

cite42_trends takes a term, a timeframe (past_7_days, past_30_days, past_90_days, past_12_months or past_5_years, with anything else falling back to 12 months), an optional geo location name that defaults to the United States, and an optional language for the related queries. Our first request was:

{
  "term": "Google Trends",
  "geo": "United States",
  "timeframe": "past_12_months",
  "language": "en"
}

The second changed geo to United Kingdom and kept everything else fixed, so the two series differ only by market. This prompt keeps Claude to that request:

Use cite42_trends for term "Google Trends",
geo "United States", timeframe "past_12_months", language "en".

Show the returned term, number of points, date range,
trend label and the related queries relevant to developers.
Explain the label using the series.
Describe values as relative interest, never as search counts.

Check the returned term before you read the result. If you send a long phrase, the tool reduces it to a short lookup term, and the series describes that term.

The US and UK results

Both requests returned 53 weekly points from October 5, 2025 to October 4, 2026, and cost $0.05 each.

The US series is easy to read: interest climbed from 31 to its March peak, then slid through the summer to 14, the lowest point of the year. The UK series was noisier, with spikes in late December, July and August, and its last six weeks fell from 59 to 16, also the year's low.

An abridged US response:

{
  "term": "google trends",
  "trend": "falling",
  "interestOverTime": [
    { "date": "2025-10-05", "value": 31 },
    { "date": "2026-03-01", "value": 100 },
    { "date": "2026-10-04", "value": 14 }
  ],
  "risingQueries": ["google trends api", "google trends data"]
}

The excerpt keeps three of the 53 points and two of the related queries. The evidence file has both full responses under trends-us and trends-uk.

Why the UK series says rising while it falls

Cite42's label compares the average of the first quarter of points with the average of the last quarter. More than 10% higher is rising, more than 10% lower is falling, and anything in between is stable. In the UK, the last 13 weeks still included the August spike, so their average (48.5) beat the first 13 weeks (38.7) by 25%, even though the series ended at its lowest value.

The label summarizes the whole window, and it can disagree with the last few weeks. Ask Claude to report both:

Read the whole returned series.
Compare the first-quarter and last-quarter averages,
then describe the final six points separately.
Report the maximum value and its date.
Explain any conflict between the label and recent movement,
and leave causes out unless you have other evidence.

For a decision about timing, use a shorter window such as past_90_days. For seasonality, use past_5_years and look for repeating peaks. An empty series also comes back labeled stable, because the label needs at least two points, so check the point count first and treat an error as a failed request rather than flat demand.

What the 0 to 100 values mean

Google Trends reports relative interest. Google's FAQ explains that each point is divided by total searches in that place and time, then scaled from 0 to 100, and that low-volume terms show as 0. A value of 100 is the highest point in that one series.

That's why the table above says nothing about which country searched more. The US and UK series are each scaled to their own peak, so a UK 60 and a US 40 could hide any absolute difference. Use a series for timing within one market, and a keyword volume source when you need scale. Ask Claude to name the unit in every summary: "interest peaked in March" follows from the data, while "demand reached 100 searches" is a misreading.

Filter related queries, then check search volume

The US response included useful candidates such as google trends api, google trends data and google trends youtube, mixed with general Google how-to queries like how to set up google calendar. Cite42 keeps related queries that share a word with your term, and with a two-word term like this one, every "google" query qualifies. Google's guide to related searches defines top queries as the most frequent and rising queries as the fastest growing. Cite42 returns the query strings without Google's growth percentages.

A related query tells you what to investigate, and a keyword tool tells you how many people search for it. Our separate US keyword check on October 5 reported 1,600 monthly searches for google trends api and 50 for google trends mcp. A follow-up prompt that adds volume:

From the returned related queries, select only terms relevant
to developers looking for Google Trends data.
Propose a cite42_keywords_batch request for country "us"
and wait for my approval.
Keep measured volumes separate from your own suggestions,
then recommend one article angle and explain the evidence.

A keyword batch of 2 to 50 terms costs $0.09 when any term has volume. The Claude Code SEO guide covers keyword research in more depth.

What this research costs

The two Trends requests above cost $0.10 in total on October 5, 2026. At current prices, the same two requests plus one keyword batch cost $0.19. Cite42 caches identical Trends requests for 24 hours, and a cached answer is billed like any successful call, so save the raw series and let Claude rework the analysis from it instead of re-running the tool.

Frequently asked questions

What are the best MCP servers for Google Trends data?

As of October 2026, the main options are Cite42, HasData's Google Trends MCP, SerpApi's MCP server, DataForSEO's MCP server, Trends MCP and Apify's Google Trends Scraper. HasData and SerpApi cover the most Trends chart types, and Trends MCP adds sources beyond Google. Cite42 charges $0.05 per Trends request with no subscription and includes keyword volume and AI visibility tools in the same server.

Is there an official Google Trends MCP server?

We found no MCP server published by Google as of October 2026, and Google's own Trends API is an application-only alpha. The Google Trends servers in MCP directories are run by companies such as Cite42, HasData, SerpApi and DataForSEO, or by individual developers.

How do I add Google Trends to Claude Code?

Run claude mcp add --transport http --scope user cite42 https://mcp.cite42.dev, then claude mcp login cite42 to sign in. Claude can then call cite42_trends for any term, location and timeframe. Each request costs $0.05 from a prepaid balance that never expires.

Is there a Google Trends MCP server without a subscription?

Yes. Cite42 bills each Trends request at $0.05 from a prepaid balance, with a $25.00 minimum top-up and no monthly fee. DataForSEO's MCP server also uses pay-as-you-go billing, with a $50 minimum payment. HasData, SerpApi and Trends MCP offer free monthly allowances with subscriptions above them.

Can Google Trends in Claude tell me how many people search for a term?

Google Trends data is a relative 0 to 100 index, so it shows timing and direction rather than search counts. For monthly volume, have Claude run a keyword tool as a separate step, such as Cite42's keyword batch, and keep the two numbers apart in its summary.

Try it on one term

Create a Cite42 account for $1.00 of free credit, connect https://mcp.cite42.dev and ask Claude for one term, one country and one timeframe. To call the same endpoint from code, the Google Trends API guide has Python and TypeScript examples and compares the API options.