Google Trends Comparison: compare up to 5 keywords in about 2 seconds

Get Google Trends interest over time for 1 to 5 keywords on one shared 0–100 scale, the same numbers you see in the Trends "Compare" view. Results arrive as clean JSON in about 2 seconds. It runs without a browser, so there's no waiting on page loads.

$0.02 per search ($0.014–0.018 on Apify's paid plans), whether you compare 1 keyword or 5. You aren't charged for failed searches.

What it's for

• Brand and competitor tracking: ["nike", "adidas", "puma"] in the US over 12 months (example)
• Content and SEO planning: check which topic is growing, or which keyword wording people search, before you write
• Market research: product demand by country (geo: "IT", "JP", …) (example) or on Google Shopping
• AI agents and automations: one call returns a structured answer to "is X more searched than Y?", with no scraping logic needed

Example tasks

Ready-made searches you can run in one click and then edit (all examples):

Brands, products and markets

• Track brand search interest vs competitors
• Compare brand demand in one country: phones in Italy
• Compare AI chatbot popularity on Google Trends

SEO, content and video

• Choose between SEO keyword variants with Google Trends
• Compare YouTube search trends for any topic
• Compare programming language popularity

E-commerce and demand planning

• Compare product demand on Google Shopping
• Spot seasonal demand with 5 years of Google Trends

Events, news and fast-moving topics

• Measure an event's search impact over custom dates
• Track crypto search interest over the past 7 days

Why this actor

On 2026-09-24 we ran the same 8 queries through this actor and the 3 most-used Google Trends actors on Apify, with a 300 s timeout and the same scoring for all:

Actor	Useful results	Median run time	Real 3-keyword comparison
This actor	8/8	1.7 s	✅
Official Google Trends Scraper	4/8 (4 timeouts)	279 s	✅
2 other popular Trends actors	7/8 each	5–7 s	❌ (no data returned)

This was a single-day test of 8 queries per actor. Results may vary at other times.

For every fixed date range, the values match the other actors within Google's normal sampling noise. Tested queries covered worldwide and single countries, past hour to all-time ranges, a non-Latin keyword (ラーメン), a low-volume keyword, and a 3-keyword comparison.

Input

{
  "keywords": ["chatgpt", "gemini", "claude"],
  "geo": "US",
  "timeRange": "today 12-m"
}

Field	Description
keywords	1–5 terms compared on one scale. One search.
searchTerms	Alternative: several independent searches, one result row each. "tea, coffee" inside one entry compares those two. Each entry is one search.
geo	ISO country code (US, IT, JP…). Empty = worldwide.
timeRange	now 1-H, now 4-H, now 1-d, now 7-d, today 1-m, today 3-m, today 12-m (default), today 5-y, all, or a custom range "2024-01-01 2024-06-30". Lower-case now 1-h also works.
category	Google Trends category id, 0 = all.
property	"" web (default), news, images, youtube, froogle (Shopping).
language	Label language, default en-US.
maxRetries	Retries when Google rate-limits a request, with backoff and a fresh session (default 3).
proxyConfiguration	Optional. Off by default; enable it only if your runs get rate-limited.

An invalid input stops the run immediately with a clear message, for example Invalid timeRange "last week". Use one of: …. It won't time out or return an empty dataset instead.

Output

One dataset row per search:

{
  "searchTerm": "chatgpt, gemini, claude",
  "keywords": ["chatgpt", "gemini", "claude"],
  "geo": "US",
  "timeRange": "today 12-m",
  "interestOverTime_timelineData": [
    {
      "time": "1758412800",
      "formattedTime": "Sep 21 – 27, 2025",
      "formattedAxisTime": "Sep 21, 2025",
      "value": [94, 25, 5],
      "hasData": [true, true, true],
      "isPartial": false
    }
  ],
  "interestOverTime_averages": [76, 27, 18],
  "pointCount": 53,
  "fetchedMs": 358
}

value holds one number per keyword, in the order you gave them. The field names match the official Google Trends Scraper's interestOverTime_* fields, so switching needs little or no change on your side.

Pricing

Apify plan	Per successful search (1–5 keywords)
Free	$0.02
Bronze	$0.018
Silver	$0.016
Gold and above	$0.014

• Failed searches are free. The actor start is billed at Apify's standard platform fee.
• Example: comparing 5 brands every day for a month is 30 searches: $0.60 on the Free plan, $0.42 on Gold.
• Spending limits: if you set a maximum cost per run, the actor stops before starting a search it can't charge for, and tells you how many searches it finished.

Use it from code, n8n or an AI agent

API (JavaScript):

import { ApifyClient } from 'apify-client';
const client = new ApifyClient({ token: process.env.APIFY_TOKEN });
const run = await client.actor('clover_engineer/google-trends-compare').call({
  keywords: ['chatgpt', 'gemini', 'claude'], geo: 'US', timeRange: 'today 12-m',
});
const { items } = await client.dataset(run.defaultDatasetId).listItems();

Python:

from apify_client import ApifyClient
client = ApifyClient("<APIFY_TOKEN>")
run = client.actor("clover_engineer/google-trends-compare").call(
    run_input={"keywords": ["chatgpt", "gemini", "claude"], "geo": "US"})
items = client.dataset(run["defaultDatasetId"]).list_items().items

n8n / Make / Zapier: use the Apify integration and choose "Run Actor and get dataset items" with the input above.

AI agents (MCP): add this actor to the Apify MCP server. Each input field is described so an agent can fill it in from a question like "Compare interest in ChatGPT, Gemini and Claude in the US over the past year."

Limits

• Interest over time only. Regional breakdown, related queries and trending searches aren't included yet.
• Values are relative, as in Google Trends itself. 100 = the peak within your comparison. Low-volume keywords and short time ranges vary slightly between requests because of Google's sampling.
• Past-hour and past-day ranges are rolling windows, so two runs a few minutes apart cover slightly different periods.
• This actor isn't affiliated with Google. It reads publicly available Google Trends data.

How it works

1. It opens a session with Google Trends, the same way the Trends website does.
2. It asks Trends for the comparison you requested: your keywords, country, time range, category and Google property.
3. It downloads the interest-over-time series as JSON. No browser is involved, which is why a search takes about 2 seconds.
4. If Google rate-limits the request, it waits, starts a fresh session and tries again (up to maxRetries times).
5. It saves one dataset row per search and charges one search event, only after the data is saved.

FAQ

Is this a Google Trends API? Google has no official public Google Trends API. This actor gives you the same interest-over-time data as the Trends website, as structured JSON you can call from code, n8n, Make, Zapier or an AI agent.

How do I compare more than 5 keywords? Google Trends compares at most 5 terms on one scale. To compare more, put one keyword (for example your main brand) in every search as a shared reference, then rescale the other keywords' values against it.

Why don't my numbers match another run exactly? Google Trends samples its data, so low-volume keywords and short time ranges can differ by a few points between requests. Past-hour and past-day ranges also move with the clock.

Does it work with pytrends inputs? Yes. timeRange accepts the same strings as pytrends' timeframe (today 12-m, now 7-d, 2024-01-01 2024-06-30…), and geo uses the same country codes.

Do I need a proxy? Usually not. Enable proxyConfiguration only if your runs report repeated rate limits (HTTP 429).

What happens if a search fails? The run stops with a clear error message and you aren't charged for that search.