Google Trends Scraper

📈 What is Google Trends Scraper?

Google Trends Scraper extracts Google Trends data for any keyword: interest over time, interest by country, state, metro area or city, top and rising related queries, plus ready-made figures (average, peak and its date, latest value, trend direction). It also compares keywords on one scale — more than 5 of them — and lists the searches trending now in any country, with their search volume, growth, start and end time, and news articles. It works as a Google Trends API alternative: run it on a schedule, call it from your code, or plug it into Make, Zapier or n8n.

Type a keyword (python), pick a country and a time range, click Start, and download the report in JSON, CSV or Excel. No login, nothing to set up. Several keywords and countries in one run, one report each — a comparison of 9 keywords takes 10 to 40 seconds, and a whole Trending Now list (300 to 2,000 searches) comes in a few seconds.

📋 What data can you extract from Google Trends?

Three kinds of results in one dataset (recordType): a keyword report per keyword and location, a comparison per location, a trending search per search trending now. 54 fields:

Category	What you get
📈 Interest over time	the 0-100 curve point by point, with its dates and which point is still unfinished
🧮 Ready-made figures	average, peak and when it happened, lowest, latest value, rising / falling / stable and by how much
🌍 Interest by region	countries, states, metro areas or cities with their score, the top region named, cities with coordinates
🔎 Related queries	the 25 top and 25 rising queries with their score or growth, "Breakout" flagged
🏷️ What was asked	keyword, Google topic, location, time range, category, search type, language
🆚 Comparisons	up to 50 keywords on one scale: each one's average, peak, latest value and share of the total, the most searched one, which keyword leads in each region
🔥 Trending Now	the searches trending in a country over the last 4 hours to 7 days: rank, search volume (200K+), growth %, start and end time, still trending or not, related searches, categories
📰 News articles (option)	the articles Google shows for each trending search: title, link, source, date, image

Every field, with an example, is listed in the Output section below.

✅ Why use Google Trends Scraper?

• 📊 Figures included: average, peak, latest value and trend direction are computed for you, the unfinished last point of a curve never skews them.
• 🆚 More than 5 keywords on one scale: Google compares 5 at most; the Actor puts up to 50 on the same scale, so you can rank a whole list of brands or products.
• 🔥 Trending Now with figures: search volume, growth, how long each trend has lasted, its categories and — as an option — its news articles, for any country, filtered by category, volume, date or keyword.
• 🧩 Several keywords and countries in one run: keywords × locations, one report each, with a cap per source — and keyword reports, comparisons and Trending Now can all go in the same run.
• 🌐 Region detail: countries, states, US metro areas or cities with their coordinates; regions Google has too little data for are left out, or kept with an empty value — never a false 0.
• 🧠 Google topics: look a keyword up as a Google topic to count every way people search for it.
• 🔔 Monitoring built in: tick Only new results, schedule the Actor, and each run returns (and charges) only what has moved since the last one.
• 🛡️ Reliable at volume: a request Google turns away is retried at once on a fresh proxy session, so a long list of keywords goes through.
• 🔌 API, scheduling, integrations (Make, Zapier, n8n, Google Sheets…), and JSON/CSV/Excel export via the Apify platform.

🚀 How to scrape Google Trends

1. Create a free Apify account.
2. Open Google Trends Scraper and type a Keyword (e.g. python), or paste your own Google Trends URLs into Google Trends URLs.
3. Pick a Country or region, a Time range, a Category and a Search type if you need them; tick Compare the keywords on one scale to put them on one scale.
4. For the searches trending now, add country codes to Trending Now countries (US, GB) and pick the period.
5. Set Max results (100 by default, 0 = no limit) — and Max results per source when you run several keywords — then click Start.
6. Download the dataset in JSON, CSV, Excel or via API.

💰 How much does it cost to scrape Google Trends?

This Actor uses pay per event pricing: you pay for each result delivered plus a small fee for each run start.

What you get	Price per 1,000 (Free plan → Gold)
Keyword report: one keyword (or one category) in one location, with its curve, regions, related queries and figures	$0.63 → $0.27
Compared keyword: each keyword of a comparison row (a comparison of 5 keywords = 5 compared keywords)	$4.49 → $4.46
Trending search: one search of a Trending Now list	$0.89 → $0.19
+ Trending search news (option, off by default): added to a trending search delivered WITH at least one news article; a trend without articles is not charged for them	+$0.45 → +$0.42
Run start: once per run	$0.01

Examples on the Free plan: 100 keyword reports cost 100 × $0.00063 + $0.01 = $0.07; a Trending Now list of 1,000 searches for one country costs 1,000 × $0.00089 + $0.01 = $0.90 (with a Gold subscription: $0.04 and $0.20).

A keyword with no data at all is not saved and not charged; a trending search dropped by a filter costs nothing. Platform usage (compute, proxy) is included in the price.

⚙️ Input

{
    "query": "python",
    "location": "US",
    "timeRange": "today 12-m",
    "maxItems": 10
}

Several keywords and countries, a custom period, only what is new since the last run:

{
    "searchQueries": ["python", "java", "rust"],
    "locations": ["US", "GB", "FR"],
    "timeRange": "custom",
    "customTimeRange": "2024-01-01 2024-06-30",
    "maxItemsPerQuery": 1,
    "onlyNew": true,
    "stateKey": "languages"
}

Nine keywords on one scale (the first one is the anchor of every group of 5, pick a popular one):

{
    "searchQueries": ["python", "java", "javascript", "rust", "golang", "kotlin", "swift", "ruby", "php"],
    "location": "US",
    "compareKeywords": true
}

The searches trending now in two countries, sports only, still trending, with their news articles:

{
    "trendingGeos": ["US", "GB"],
    "trendingHours": "24",
    "trendingCategories": ["17"],
    "trendingActiveOnly": true,
    "includeTrendingNews": true,
    "maxItemsPerQuery": 20
}

Or with your own Google Trends pages:

{
    "startUrls": [
        { "url": "https://trends.google.com/trends/explore?q=python,java&geo=US&date=today%205-y" },
        { "url": "https://trends.google.com/trends/explore?cat=71&geo=FR&date=now%207-d&gprop=youtube" }
    ],
    "maxItems": 20
}

Field	Notes
query, searchQueries	A keyword exactly as typed in Google (python), or a Google topic id (/m/05z1_); searchQueries adds more keywords, one report each.
location, locations	Google location code: US, GB, FR, a state (US-CA), a US metro area (US-CA-807). Empty = worldwide. Every keyword is analysed in every location (max 2000 searches per run).
startUrls	Google Trends pages: an Explore page (https://trends.google.com/trends/explore?q=python&geo=US) — its keywords, country, time range, category and search type are used as they are, and a page of several keywords (q=python,java) gives one comparison (one place per page: Google's comparison across places, q=python,java&geo=US,FR, is refused) — or a Trending Now page (https://trends.google.com/trending?geo=US&hours=24), its category and its "active only" setting kept.
timeRange	Period of the curve: now 1-H, now 4-H, now 1-d, now 7-d, today 1-m, today 3-m, today 12-m, today 5-y, all (2004 to now) or custom.
customTimeRange	With a custom time range: two days (2024-01-01 2024-06-30) or two hours, under 8 days (2025-03-01T10 2025-03-03T10).
category	Google category id, the cat value in the address of an Explore page: 0 is all categories, 71 Food & Drink, 5 Computers & Electronics. The result names it. With a category and no keyword you get the category's own curve and related queries.
property	Which Google search the data comes from: images, news, froogle (Shopping) or youtube; empty is the web.
resolveTopics	Look each keyword up as a Google topic and analyse the topic: it counts all the ways people search for it, in every language. The topic found is in the result. Not applied to a comparison: its keywords are compared as typed.
timezoneOffsetMinutes	Minutes west of UTC as Google expects it: 0 is UTC, 300 is New York, -60 is Paris. Only the formattedTime text of each point follows it: the points, their days and every date in the results stay in UTC.
language	Language of the region names and queries: en-US, fr, de, es.
compareKeywords	Compare the keywords of a location on one common scale instead of one report each: one comparison row per location, up to 50 keywords. With a single keyword you get its normal report.
includeInterestOverTime, includeInterestByRegion, includeRelatedQueries	What each report holds (all on by default; at least one must stay on). A comparison always holds its curve.
regionResolution	Level of the region list: auto (the level Google gives the place: countries for the world, regions inside a country, metro areas inside a US state), COUNTRY, REGION, DMA (US metro areas) or CITY. A level you pick is sent as it is, and Google refuses the ones a place has none of (countries inside a country, regions inside a state).
includeLowVolumeRegions	Keep the regions Google has too little data for, with an empty value (never 0).
trendingGeos	Country codes (US, GB, FR): each gives its Trending Now list. Empty = no Trending Now.
trendingHours	Trends started in the last 4, 24, 48 or 168 hours.
trendingCategories	Keep only these Trending Now categories (17 is Sports, 4 Entertainment). Empty = all.
trendingActiveOnly, includeTrendingNews	Only the searches still trending; the news articles Google shows for each trending search (paid option, off by default).
minSearchVolume	Keep only the trending searches with at least this many searches: 200, 1000, 10000.
maxItems	Stop after this many results for the whole run (0 = unlimited); each keyword of a comparison counts as one result.
maxItemsPerQuery	Cap for EACH source (a keyword in a location, a comparison, a Trending Now country); each keyword of a comparison counts as one result. 0 = no cap.
postedAfter, postedBefore	Trending Now only: the searches that started after / before this date: 2026-10-01, or a period before now (6 hours, 2 days).
excludeKeywords	Trending Now only: drop the searches whose title contains one of these words (case and accents ignored).
stateKey, onlyNew, resetState	Monitoring: only what a run with this memory key has not delivered (us-trending); resetState forgets the memory.
proxyConfiguration, requestsPerSession, maxConcurrency, maxRequestsPerMinute, minRequestIntervalMs, maxRequestRetries, debugLog	Advanced: the Apify proxy by default (included in the price; the residential proxy is not available), requests per IP address, speed and retries.

📦 Output

One real result, shortened (the curve and the lists keep their first entries only):

{
    "id": "kw:python|US|today 12-m|0|web",
    "recordType": "keyword",
    "url": "https://trends.google.com/trends/explore?q=python&geo=US&date=today%2012-m",
    "keyword": "python",
    "topicId": null,
    "geo": "US",
    "geoName": "United States",
    "timeRange": "today 12-m",
    "startDate": "2025-10-05T00:00:00.000Z",
    "endDate": "2026-10-04T00:00:00.000Z",
    "resolution": "WEEK",
    "category": 0,
    "categoryName": "All categories",
    "property": "web",
    "averageInterest": 60.1,
    "peakInterest": 100,
    "peakDate": "2026-06-14T00:00:00.000Z",
    "latestInterest": 39,
    "latestIsPartial": true,
    "trendDirection": "falling",
    "trendChangePercent": -18.6,
    "minInterest": 37,
    "topRegion": "Wyoming",
    "topRegionCode": "US-WY",
    "topRegionValue": 100,
    "topRelatedQuery": "what is python",
    "topRisingQuery": "walmart near me",
    "interestOverTime": [{ "date": "2025-10-05T00:00:00.000Z", "time": 1759622400, "formattedTime": "Oct 5 – 11, 2025", "value": 59, "isPartial": false }],
    "interestByRegion": [{ "geoCode": "US-WY", "geoName": "Wyoming", "value": 100 }],
    "regionResolution": "REGION",
    "relatedQueriesTop": [{ "query": "what is python", "value": 100, "formattedValue": "100", "link": "https://trends.google.com/trends/explore?q=what+is+python&date=today+12-m&geo=US" }],
    "relatedQueriesRising": [{ "query": "walmart near me", "growthPercent": null, "formattedValue": "Breakout", "isBreakout": true, "link": "https://trends.google.com/trends/explore?q=walmart+near+me&date=today+12-m&geo=US" }],
    "language": "en-US",
    "source": "python",
    "scrapedAt": "2026-10-06T00:24:44.507Z"
}

A comparison row (real, shortened: the keys left out are the same as above, empty for a comparison):

{
    "id": "cmp:python~java~javascript~rust~golang|US|today 12-m|0|web",
    "recordType": "comparison",
    "keyword": "python vs java vs javascript vs rust vs golang",
    "keywords": ["python", "java", "javascript", "rust", "golang"],
    "winner": "python",
    "anchorKeyword": null,
    "keywordStats": [{ "keyword": "python", "average": 60, "peak": 100, "peakDate": "2026-06-14T00:00:00.000Z", "latest": 38, "sharePercent": 46.2, "regionsLed": 46 }],
    "interestOverTime": [{ "date": "2025-10-05T00:00:00.000Z", "time": 1759622400, "formattedTime": "Oct 5 – 11, 2025", "values": [59, 28, 9, 25, 2], "isPartial": false }],
    "interestByRegion": [{ "geoCode": "US-DC", "geoName": "District of Columbia", "values": [62, 17, 7, 13, 1], "leader": "python" }]
}

A trending search with its news articles (real, shortened):

{
    "id": "trend:US|robert kelker kelly|2026-10-05T17:50:00.000Z",
    "recordType": "trending",
    "title": "robert kelker kelly",
    "geo": "US",
    "rank": 1,
    "searchVolume": 200000,
    "formattedSearchVolume": "200K+",
    "growthPercent": 1000,
    "startedAt": "2026-10-05T17:50:00.000Z",
    "endedAt": null,
    "isActive": true,
    "durationMinutes": 775,
    "trendBreakdown": ["robert kelker kelly", "days of our lives"],
    "categories": [{ "id": 4, "name": "Entertainment" }],
    "newsArticles": [{ "title": "Robert Kelker-Kelly Dies: ‘Days Of Our Lives’, ‘Another World’ Actor Was 62", "url": "https://deadline.com/2026/10/robert-kelker-kelly-dead-days-of-our-lives-another-world-1237146925/", "source": "Deadline", "publishedAt": "2026-10-05T19:32:00.000Z", "imageUrl": "https://encrypted-tbn1.gstatic.com/images?q=tbn:ANd9GcTJ4j9lG98oDtHhQ_5I3f3LGBvusYedVAX7f04Vqtb_HD2gQyXryDNP2o_VXpg" }]
}

You can download the dataset in various formats such as JSON, HTML, CSV or Excel.

All 54 fields

Fields	What you get
id, recordType, url	the result's stable id, its kind (keyword, comparison or trending), the same view on trends.google.com
keyword, topicId, topicTitle, topicType, topicSuggestions	the keyword analysed; the Google topic when the keyword is one or was looked up, with every topic suggested
geo, geoName, timeRange, startDate, endDate, resolution, category, categoryName, property, language, source	where, when and what kind of search; the first and last point of the data; the step of the curve; the input it came from
averageInterest, peakInterest, peakDate, latestInterest, latestIsPartial, minInterest	the figures of the curve (the unfinished last point left out), and whether the last point is unfinished
trendDirection, trendChangePercent	rising, falling or stable: the last quarter of the curve against the first, within ±10 %
interestOverTime	one point per step: date, time, label, value 0-100 (empty when Google has no data), unfinished or not
interestByRegion, regionResolution, topRegion, topRegionCode, topRegionValue	the regions with their 0-100 score, best first; cities carry coordinates instead of a code
relatedQueriesTop, relatedQueriesRising, topRelatedQuery, topRisingQuery	the related queries with their score or growth in %; a "Breakout" has no number and isBreakout set
keywords, keywordStats, winner, anchorKeyword	comparison results: the keywords, their figures, the most searched one, the anchor of a comparison of more than 5
title, rank, searchVolume, formattedSearchVolume, growthPercent, startedAt, endedAt, isActive, durationMinutes, trendBreakdown, categories, newsArticles	Trending Now results: the search, its place in Google's list, its volume and growth, when it started and ended, its related searches and categories, and its news articles
scrapedAt	ISO timestamp of the extraction

💡 Tips

How to get more results

Add keywords to searchQueries and countries to locations: each keyword is analysed in each country. Google compares the interest of a keyword with its own highest point (100), so two reports are not on the same scale: use the comparison option to put several keywords on one.

How to reduce costs

The price is per result, so the levers are maxItems, maxItemsPerQuery, and onlyNew for recurring runs (a report whose last data point has not moved, or a trending search already delivered, is skipped and not charged). For Trending Now, the filters (trendingCategories, minSearchVolume, trendingActiveOnly, dates, excludeKeywords) drop searches for free, before any charge and before any news article is fetched. Leave out the map and the related queries (includeInterestByRegion, includeRelatedQueries) when you only need the curve: the run is faster.

Compare more than 5 keywords

Google compares 5 keywords at most. Tick compareKeywords with more: the Actor reads them in groups of 5 that all share the first keyword (the anchor), then puts every group on the anchor's scale, so the highest point of the whole row is 100. Put a popular keyword first: if Google has no data for the anchor in a group, that comparison is not made (the run says why) rather than given on a false scale; if the anchor is very small next to the others, the row comes with a warning in the log. On the ranges of the last hours (now 1-H, now 4-H, now 1-d) Google sets the window of each group at the minute it is read: the row keeps the time steps every group has (a step or two fewer at the edges, at most). With 5 keywords or fewer the averages are Google's own. In the region map of a comparison, the values are each keyword's share of the searches in that region (they add up to 100), and leader names the keyword ahead.

Trending Now and news articles

Each country of trendingGeos gives its whole Trending Now list for the period (about 80 searches in 4 hours and 2,000 in 7 days for the US). rank is the search's place in Google's list before any filter. searchVolume is the lower bound of Google's bracket (200K+ → 200000). With includeTrendingNews, the news articles are fetched only for the searches you keep, after the filters and the caps.

Which time range and step

Google picks the step of the curve from the range: minutes for the last hours, hours for 7 days, days up to 9 months, weeks up to 5 years, months for 2004 to now. The last point is often unfinished (isPartial): it stays in the curve but never enters the figures.

Monitoring: only the new results

Tick Only new results (onlyNew) and schedule the Actor. The first run returns everything; each later run skips the reports whose last data point has not moved since the one delivered, and the trending searches already delivered (same country, search and start time): they are not saved and not charged. The memory lives in a named key-value store of your account (google-trends-scraper-seen, up to 150,000 results per key) and is only updated with results that really reached the dataset, so a failed run never hides anything. Give each schedule its own stateKey, and tick resetState once to start over. A trending search that does not pass a time-dependent filter yet (a relative "posted before" date, a minimum search volume) is checked again on the next run.

🔌 Integrations and API

Call the Actor via the Apify API, the JavaScript or Python clients, or connect it with integrations and webhooks (Make, Zapier, n8n, Google Sheets, Slack, Airtable…). The dataset can be fetched as JSON or CSV from any tool.

🤖 Use with AI agents (MCP)

AI agents (Claude, ChatGPT, Cursor…) can find and run this Actor through the Apify MCP server, billed to their Apify account like any run. It returns one result per keyword and country, with the interest curve, the regions, the related queries and ready-made figures; a comparison of several keywords on one scale (compareKeywords); or the searches trending now in a country (trendingGeos). Actor id: nice_dev/google-trends-scraper; MCP server with this Actor only: https://mcp.apify.com/?tools=fetch-actor-details,nice_dev/google-trends-scraper.

Smallest input, for a cheap first call:

{
    "query": "python",
    "location": "US",
    "maxItems": 1
}

Key output fields: keyword, geo, averageInterest, peakInterest, latestInterest, trendDirection, topRegion, topRelatedQuery and topRisingQuery; the full curve, regions and queries are in interestOverTime, interestByRegion, relatedQueriesTop and relatedQueriesRising. A comparison adds keywords, keywordStats and winner; a trending search has title, rank, searchVolume, growthPercent, startedAt, isActive and, with the option, newsArticles.

Cost: pay per event, see the pricing section above. Cap each call with maxItems and, through the API, with the run option maxTotalChargeUsd.

❓ FAQ

Is it legal to scrape Google Trends?

The Actor only reads what Google Trends shows publicly to any anonymous visitor: aggregated search interest and public news headlines. It logs in to nothing, and the results hold no personal data. You are responsible for using the data in compliance with Google's Terms of Service and applicable law. This Actor is not affiliated with Google.

Does it need a login or a proxy?

No login. The proxy is included in the price: leave the default setting (the residential proxy is not available). A request Google turns away is retried at once on a new proxy session, up to 10 times on top of the retries (without a proxy, after a pause of 5 seconds, doubled at each retry up to 300 seconds).

What does a result with "no data" mean?

When Google has no data at all for a keyword in that place and period (too few searches), the Actor saves nothing for it and does not charge it: the run's last line counts these reports. Inside a report, a point, a region or a query Google has no value for is empty (null), never 0.

Known limitations

• Values are relative (0-100, the highest point of the report is 100): two reports are not on one scale unless you compare the keywords in one run.
• Google gives the 25 top and 25 rising related queries, up to 250 countries, 51 US states, 210 US metro areas and 200 cities per map; cities with too little data have no value.
• Trending Now keeps the last 7 days at most, and Google has no Trending Now list for some countries (Luxembourg, Iceland…): such a country is said in the log and skipped, the other sources are delivered.
• With news articles, the trending searches are saved as their articles arrive, not in Google's order: sort on rank to get it back.
• A comparison is limited to 50 keywords per location; its first keyword must have data in every group of 5.
• Two runs sharing the same stateKey at the same time may both return the same new result.

A run that reaches its timeout stops itself about 45 seconds before it: no new report is asked, what it read is saved and, with onlyNew, remembered, and the run ends Succeeded with "Stopped before the run's timeout". Resurrect it to go on from there, or give the next run a longer timeout (Run options).

A run the platform stops without warning (out of memory)

• Resurrect it: it goes on from where it stood at most a minute before the stop. What it had read since is read again, and the results already saved are skipped — even a report whose data gained a point in between: none is delivered or charged twice, and maxItems still counts them (a comparison for each of its keywords).
• With onlyNew, the memory is saved within 2 seconds of each write: resurrect the stopped run and the results it had saved meanwhile join the memory; leave it stopped for good, and the next run may return the results of its last 2 seconds once more.

Something doesn't work?

The last line of the log counts the results saved, the reports Google had no data for, the trending searches whose news articles could not be read (delivered without them, and not charged for them), and the requests that failed after every retry. Those requests are listed, with the reason, in the FAILED_REQUESTS record of the run's key-value store. A run that saved nothing and had failed requests fails, and its last message gives the cause. If Google changes the shape of its answers, you are told instead of paying for blank rows: the report fails with the reason and is listed in FAILED_REQUESTS.

🛟 Support

Open an issue in the Issues tab with a link to your run: the run log and the FAILED_REQUESTS record of the key-value store show exactly which keywords failed and why.