Google Trends Scraper – Compare up to 2,000 Keywords

Get Google Trends data for a whole keyword list in one run: interest over time, interest by region, top and rising related queries, and a ready-made trend verdict (rising / stable / falling) for every term. Works for Web search, YouTube, Google News, Google Images and Google Shopping, any country or region, any period from the past hour to 2004.

Google Trends only lets you compare 5 terms at a time. This Actor can put 10, 50 or 500 terms on one common 0–100 scale and rank them, which Google Trends itself doesn't do.

What you get

Data	Details
Interest over time	Every data point Google shows (hourly, daily, weekly or monthly, depending on the period), with partial points flagged
Summary numbers	Average, latest, peak value and peak date, % change (latest quarter vs earliest quarter of the period), trend direction
Interest by region	Countries, regions/states, cities or US metro areas (DMA), sorted by interest
Related queries	Top queries and rising queries with their growth (+250%, Breakout) – ideal for keyword research
Common-scale comparison	Up to 2,000 terms on one scale, plus a popularity ranking
Trending now	The full list from trends.google.com/trending for any country (often 300+ trends a day in large countries): search volume, % increase, start/end time, active or not, categories and the searches that make up each trend. Filter by past 4 h / 24 h / 48 h / 7 days and by category
Search types	Web, YouTube, News, Images, Shopping
Link	A trendsUrl that opens the same view on trends.google.com, so you can check any number yourself

Why this Actor

• Built for Google's rate limits. Google Trends answers bursts of requests with HTTP 429. Each parallel worker has its own proxy session and cookie and switches to a fresh IP when Google blocks it, so runs keep going instead of stalling.
• No hanging runs. The Actor watches the run timeout and stops cleanly before it, saving everything it has already collected. You are only charged for terms that were delivered.
• Verified against Google Trends. The interest-over-time values match the data trends.google.com shows in the browser, point for point (we check this with a real browser, see How accurate is it?).
• Spreadsheet-friendly. Choose one row per term (with summary columns) or one row per term and date to chart it straight away in Excel or Google Sheets.
• Honest output. When Google has no data for part of a request, the term gets a warnings field that explains what is missing.

How to use it

1. Enter your search terms, one per line.
2. Choose a location (US, GB, ES, US-CA, or empty for worldwide), a time range and the search type.
3. Optional: turn on Compare all terms on one scale.
4. Click Start. Download the results as JSON, CSV, Excel or HTML, or call the Actor from the API, Make, Zapier, n8n or an AI agent (MCP).

Example input

{
  "searchTerms": ["chatgpt", "gemini", "claude", "perplexity", "copilot", "deepseek", "grok"],
  "geo": "US",
  "timeRange": "today 12-m",
  "property": "web",
  "compareOnSameScale": true
}

Example output (one term, shortened)

{
  "searchTerm": "bitcoin",
  "geo": "ES",
  "timeRange": "today 3-m",
  "periodResolved": "2026-06-26 2026-09-26",
  "dataResolution": "DAY",
  "property": "web",
  "averageInterest": 49.83,
  "latestInterest": 48,
  "peakInterest": 100,
  "peakDate": "2026-08-21T00:00:00.000Z",
  "trendChangePercent": -6.6,
  "trendDirection": "stable",
  "topRegion": "La Rioja",
  "topRelatedQuery": "precio bitcoin",
  "topRisingQuery": "how to buy bitcoin safely",
  "interestOverTime": [
    { "date": "2026-06-26T00:00:00.000Z", "formattedTime": "Jun 26, 2026", "value": 70, "hasData": true, "isPartial": false }
  ],
  "interestByRegion": [
    { "geoCode": "ES-RI", "geoName": "La Rioja", "value": 100 },
    { "geoCode": "ES-NC", "geoName": "Navarre", "value": 82 }
  ],
  "relatedQueriesTop": [
    { "query": "precio bitcoin", "value": 100, "formattedValue": "100", "isBreakout": false, "link": "https://trends.google.com/trends/explore?q=precio+bitcoin&date=today+3-m&geo=ES" }
  ],
  "relatedQueriesRising": [
    { "query": "how to buy bitcoin safely", "value": 700, "formattedValue": "+700%", "isBreakout": false, "link": "https://trends.google.com/trends/explore?q=how+to+buy+bitcoin+safely&date=today+3-m&geo=ES" }
  ],
  "trendsUrl": "https://trends.google.com/trends/explore?date=today+3-m&geo=ES&q=bitcoin&hl=en-US"
}

Example Trending now item

{
  "geo": "US",
  "rank": 1,
  "title": "oklahoma vs georgia",
  "searchVolume": 500000,
  "searchVolumeLabel": "500K+",
  "increasePercent": 1000,
  "startedAt": "2026-09-26T18:40:00.000Z",
  "endedAt": null,
  "isActive": true,
  "categories": ["Sports"],
  "trendBreakdown": ["ole miss vs florida", "texas tennessee", "florida football", "…"],
  "trendsUrl": "https://trends.google.com/trends/explore?q=oklahoma%20vs%20georgia&date=now%201-d&geo=US"
}

In common-scale mode each term also has a commonScale object (anchorTerm, scaleFactor, rank), and each timeline point keeps Google's original value in rawValue. The ranking is also saved in the RUN_SUMMARY record of the key-value store.

Pricing

Pay only for results:

Event	Price
One search term analyzed (all selected data for that term)	$0.002 ($2 per 1,000 terms)
One Trending now search	$0.0005 ($0.50 per 1,000)

Terms that fail or are skipped are not charged. Set a maximum cost per run in the run options and the Actor stops cleanly when it is reached.

How the common scale works

Google normalizes every request so the highest point is 100, which is why numbers from different requests can't be compared. The Actor splits your list into groups of 5 that all share one anchor term, then rescales each group so the anchor has the same values everywhere, and finally rescales everything so the highest point of all terms is 100.

• For up to 5 terms this is exactly Google's own comparison.
• Pick an anchor of medium-to-high popularity (option Anchor term). If the anchor is tiny next to a much bigger term in the same group, Google rounds it to 0–1 and precision drops. The Actor flags those terms with a Low precision warning.
• Google rounds each value to a whole number, so terms that are 1,000 times smaller than the biggest term will show values near 0. That's a limit of Google Trends itself.

FAQ

How accurate is it? The Actor reads the same JSON that the Google Trends website loads. We compared a 93-day series point by point with the values trends.google.com served to a real browser for the same search, and all 93 matched. Google Trends is based on a sample of searches, so values for small terms can change slightly between days. This is how Google Trends works, not an error.

Do I need a proxy? Google rate-limits datacenter IPs quickly, so the default is Apify's residential proxy. If your account doesn't have it, the Actor falls back to the default proxy automatically.

Why are related topics empty? Google currently returns empty related topics lists for most searches, including on its own website, so that option is off by default. Related queries work normally.

What do the values mean? 100 is the peak popularity for the term in the chosen place and period, 50 means half as popular, and 0 means not enough data. They are relative numbers, not search volumes.

Can I use it from an AI agent? Yes. Call it through the Apify MCP server or the API. The summary fields (trendDirection, trendChangePercent, topRisingQuery) are designed to be read by LLMs without processing the full timeline.

Is Trending now the same list as on the website? Yes. It reads the same data that trends.google.com/trending loads, and the category filter works the same way (we checked that the number of trends per category matches the website's menu for all 20 categories). If that call ever fails, the Actor falls back to Google's short RSS feed (about 10 trends per country) and logs a warning.

What are the limits? Up to 2,000 terms per run. Each term needs 2–5 requests to Google (depending on the data you choose), and Google slows down clients that ask too fast, so for lists of several hundred terms raise the run timeout or split the list. If the timeout gets close, the Actor stops cleanly and keeps what it has.

Legal

This Actor only reads public, aggregated and anonymous data from Google Trends and the public Trending Now RSS feed. It doesn't collect any personal data. It isn't affiliated with or endorsed by Google. Check that your use complies with Google's terms and the law where you live.