Google Trends Scraper: Related Queries & Trending Searches

Get Google Trends data as clean JSON, the same numbers you see on trends.google.com, for marketers, SEO teams and analysts who track demand over time.

Use cases, each with a ready-to-run example:

• Compare keywords over 12 months, such as ChatGPT vs Claude vs Gemini
• Get today's trending searches in the US, the UK and India
• Find the rising related queries for a keyword
• See interest by US metro area
• Compare YouTube search trends
• Compare Google Shopping trends

Compare up to 5 keywords at once, for any country or region, time range, category and search type, or up to 50 on one scale. Paste a trends.google.com link, break interest down by city or US metro area, or research a keyword's related searches in one call. You can also pull the searches trending right now in any country.

You pay $1 per 1,000 keyword comparisons ($0.001 each), with up to 5 keywords in each one. Trending Now costs $0.25 per 1,000 trending searches, and less on Apify Store discount tiers (Bronze, Silver, Gold), down to $0.175 per 1,000 on Gold. There is no start fee, and comparisons that fail or have no data are never charged.

At a glance

Topic	Details
Input	Comparisons of 1 to 5 keywords each (up to 50 with moreThanFiveKeywords), with country or region, time range, category and search type. trends.google.com explore and trending links. Research keywords. Optional country codes for Trending Now
Output	JSON rows with interest over time, interest by region (countries, subregions such as US states, cities or US metro areas) and top and rising related queries per keyword. Each researched related query's own timeline and growth. Trending Now rows with search volume and start time
Price	$0.001 per comparison ($1 per 1,000), with its timeline, region and related-query rows included. Trending Now is $0.00025 per search ($0.25 per 1,000), down to $0.175 per 1,000 on Gold. No start fee. Failed or empty comparisons are free
Tested	Latest test run: 58 of 58 inputs returned data, with 15 comparisons of 6 to 13 keywords, 11 city or metro breakdowns, 18 trends.google.com links and 12 research keywords. The run before it: 58 of 58 comparisons and 7 of 7 Trending Now countries
Limits	Google compares at most 5 keywords, so more than 5 are fetched in batches. No related topics. Cities only where Google has enough data

What you get

For each comparison:

Row type	What it holds
interestOverTime	The timeline for all keywords on the same 0 to 100 scale, plus the average per keyword
interestByRegion	Countries (for worldwide) or subregions such as US states (for a country), with a value per keyword. Cities or US metro areas with regionResolution
relatedQueries	One row per keyword with the top and rising related searches, including Breakout flags

For each research keyword, its comparison rows above, then:

Row type	What it holds
relatedTrend	One related query's own timeline, its average and its growth over the last quarter of the time range
researchSummary	Free. Every researched query with its growth, highest growth first

For each Trending Now country, one row per trending search with search volume, percent increase, start and end time, categories and the related searches Google groups with it. Its trendId joins the country, the search and its start time, so the same trend keeps the same ID across runs.

Every row has a status of ok, no_data, invalid_input or error, and a message when it is not ok. Rows from a pasted link carry it as inputUrl.

Input

Type comparisons one per line, with keywords separated by commas, like the compare box on Google Trends.

{
  "comparisons": ["chatgpt, gemini, claude", "bitcoin"],
  "geo": "US",
  "timeRange": "today 3-m",
  "property": "web",
  "trendingCountries": ["US", "GB"]
}

Setting	Values
geo	Empty for worldwide, a country such as US or IN, or a region such as US-CA or GB-ENG
timeRange	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, or custom with customTimeRange such as 2024-01-01 2024-06-30
category	Google Trends category id, 0 for all. It is the cat= value in a trends.google.com URL
property	web, youtube, news, images or shopping

Through the API, any comparison can be an object that overrides these settings for that comparison only:

{
  "comparisons": [
    {"keywords": ["tea", "coffee"], "geo": "GB", "timeRange": "all"},
    {"keywords": "iphone, samsung galaxy, google pixel", "geo": "IN", "property": "youtube"},
    {"keywords": ["black friday"], "geo": "US", "timeRange": "custom", "customTimeRange": "2023-10-01 2023-12-31", "property": "shopping"}
  ]
}

More options

All of these are off until you set them, so a run without them works exactly as before.

Setting	What it does
regionResolution	city gives cities, metro gives US metro areas (Nielsen DMA codes), default keeps countries or subregions. It can also go in a comparison object
moreThanFiveKeywords	Lets one comparison hold 6 to 50 keywords on one 0 to 100 scale
trendsUrls	Links copied from trends.google.com. An explore link runs that comparison with its own location, time range, category and search type. A trending link runs Trending Now for its country, window, category and active filter
researchKeywords	Seed keywords. Each one returns its comparison and related queries, then the own timeline and growth of each of its top related queries
researchMaxRelated	How many related queries get their own timeline per seed, 10 by default and 25 at most
researchRelatedType	both (rising first, then top), rising or top
onlyNewSinceLastRun	Trending Now only. Returns only the trending searches this scraper has not returned to you before for the same country and categories. See Run it on a schedule

{
  "comparisons": ["chatgpt, gemini, claude, perplexity, copilot, deepseek, grok, mistral"],
  "moreThanFiveKeywords": true,
  "regionResolution": "metro",
  "geo": "US",
  "trendsUrls": ["https://trends.google.com/trends/explore?cat=7&date=today%2012-m&geo=US&q=mortgage%20rates,inflation"],
  "researchKeywords": ["pickleball"],
  "researchMaxRelated": 5
}

More than 5 keywords on one scale

Google compares at most 5 keywords at a time. With moreThanFiveKeywords, the scraper sends the first 5 as one comparison, then takes the most searched keyword so far as the anchor and sends it with the next 4, and so on. The anchor's values in both comparisons give the factor that puts the new keywords on the first scale, and the result is scaled so the top value is 100. Regions are joined the same way, through the anchor's share in each region. A region where a later batch has no data for the anchor is left out, and the row's message says how many. The row lists batchCount, anchorKeywords and the Google link of each batch in batchExploreUrls.

Run hKjd0PGLalXj5a1Vv measured how close it gets. It checked 13 keyword pairs from 3 merged comparisons, each against a direct 2-keyword comparison of the same pair. The ratio between the two keywords differed by a median of 0.9%, and by 3.6% or less for 11 of the 13. The other 2 were kotlin against java (23%) and scala against javascript (15%), where the smaller keyword sits at 1 on the merged scale and Google's rounding to whole numbers dominates.

City and metro areas

Google shows cities only where it has enough data, the same as on the website. In the latest test run, 4 of 6 city breakdowns returned cities (2 to 20 each), and tacos against burritos in Texas and oktoberfest in Germany over 5 years returned none. Metro areas exist only for US locations. All 4 US metro breakdowns returned data, 210 areas for the whole US and 10 for New York state. A metro breakdown for Australia returns a free no_data region row that says so.

Research from a seed keyword

growthPercent compares the average of the last quarter of the time range with the quarter before. For the default 12 months of weekly points, that is the last 13 weeks against the 13 before, and growthWindowPoints says how many points each side held. A query that had no searches in the quarter before has growthPercent: null and previousAverage: 0, and the summary lists it first. Google's own rising percentage (relatedValue) compares the whole time range with the one before it, so the two numbers measure different things. Rising related queries come from Google as they are, and they sometimes include searches unrelated to the seed.

Run it on a schedule

Trending Now changes within minutes, so a scheduled run with onlyNewSinceLastRun: true works as an alert for new trending searches in the countries you watch. Each run still keeps only each country's top trendingMaxItems, and returns the ones you have not had before. We ran 59 countries and 3 trending links, 20 searches each, four times on 20 Oct. The first run returned 1,223 searches. Runs 2, 15 and 49 minutes after the first returned 4, 102 and 150 searches, none of them returned before. None was a search from an earlier run under a new start time, so each trendId held. A country with nothing new comes back as a free no_data row. A daily run with the default 24-hour trendingHours window covers each day, and the 4-hour window suits a run every 4 hours.

1. Set your input with onlyNewSinceLastRun: true and click Save as a new task.
2. On the task page, click Actions, then Schedule, and pick how often it runs. Apify's schedules guide covers cron times and time zones.
3. On the task's Integrations tab, choose where each run's results go.

Integration	What it does
Slack	Posts a message to your channel when a run finishes
Google Drive	Saves each run's results as a file in your Drive
Webhook	Sends the finished run, with its dataset ID, to your own URL
Make, Zapier, n8n	Adds each new row to a Google Sheet, a CRM or a database

Each trending row has a stable trendId, so a sheet or database can skip rows it already holds. Keyword comparisons have no only-new mode, and a scheduled comparison returns the full timeline each time.

Output example

This is real output from a test run, for the comparison chatgpt, gemini, claude in the US over the past 90 days. Lists are cut to 3 entries here; the real rows have the full 93 days, 51 states and 25 top queries.

{
  "type": "interestOverTime",
  "status": "ok",
  "comparisonIndex": 1,
  "keywords": ["chatgpt", "gemini", "claude"],
  "geo": "US",
  "timeRange": "today 3-m",
  "category": 0,
  "property": "web",
  "exploreUrl": "https://trends.google.com/trends/explore?q=chatgpt%2Cgemini%2Cclaude&date=today+3-m&geo=US",
  "summary": "93 points, averages chatgpt 65, gemini 26, claude 18",
  "averages": {"chatgpt": 65, "gemini": 26, "claude": 18},
  "timeline": [
    {"date": "2026-07-09T00:00:00Z", "formattedTime": "Jul 9, 2026", "values": {"chatgpt": 70, "gemini": 25, "claude": 28}, "isPartial": false},
    {"date": "2026-07-10T00:00:00Z", "formattedTime": "Jul 10, 2026", "values": {"chatgpt": 66, "gemini": 23, "claude": 24}, "isPartial": false},
    {"date": "2026-07-11T00:00:00Z", "formattedTime": "Jul 11, 2026", "values": {"chatgpt": 48, "gemini": 20, "claude": 16}, "isPartial": false}
  ]
}

{
  "type": "interestByRegion",
  "status": "ok",
  "keywords": ["chatgpt", "gemini", "claude"],
  "geo": "US",
  "summary": "51 regions, top California",
  "resolution": "REGION",
  "valueType": "share",
  "regions": [
    {"geoCode": "US-CA", "geoName": "California", "values": {"chatgpt": 57, "gemini": 25, "claude": 18}},
    {"geoCode": "US-GA", "geoName": "Georgia", "values": {"chatgpt": 63, "gemini": 22, "claude": 15}},
    {"geoCode": "US-NJ", "geoName": "New Jersey", "values": {"chatgpt": 61, "gemini": 23, "claude": 16}}
  ]
}

{
  "type": "relatedQueries",
  "status": "ok",
  "keyword": "claude",
  "geo": "US",
  "timeRange": "today 3-m",
  "summary": "25 top, 8 rising",
  "top": [
    {"query": "claude ai", "value": 100, "link": "https://trends.google.com/trends/explore?q=claude+ai&date=today+3-m&geo=US"},
    {"query": "claude code", "value": 56, "link": "https://trends.google.com/trends/explore?q=claude+code&date=today+3-m&geo=US"},
    {"query": "what is claude", "value": 25, "link": "https://trends.google.com/trends/explore?q=what+is+claude&date=today+3-m&geo=US"}
  ],
  "rising": [
    {"query": "kimi k3", "value": 10750, "formattedValue": "Breakout", "isBreakout": true, "link": "https://trends.google.com/trends/explore?q=kimi+k3&date=today+3-m&geo=US"}
  ]
}

A Trending Now row for the UK, from run XBJhdCLvhfqzWaaK1, with related searches cut to 5:

{
  "type": "trendingNow",
  "country": "GB",
  "hours": 24,
  "status": "ok",
  "rank": 2,
  "query": "man city vs aek",
  "searchVolume": 50000,
  "increasePercent": 1000,
  "startedAt": "2026-10-20T17:50:00Z",
  "endedAt": null,
  "isActive": true,
  "categoryIds": [17],
  "categories": ["Sports"],
  "relatedQueries": ["aek athens", "elliot anderson", "aek athens f.c.", "aek", "man city – aek"],
  "exploreUrl": "https://trends.google.com/trends/explore?q=man+city+vs+aek&geo=GB&date=now+1-d",
  "trendId": "GB:man city vs aek:2026-10-20T17:50:00Z"
}

Output of the extra options

These rows come from run FN6FJIO2hIkTuX729, with lists cut to 3 entries. Eight AI assistants on one scale in the US over the past 90 days, from 2 Google comparisons through the anchor chatgpt:

{
  "type": "interestOverTime",
  "status": "ok",
  "comparisonIndex": 2,
  "keywords": ["chatgpt", "gemini", "claude", "perplexity", "copilot", "deepseek", "grok", "mistral"],
  "geo": "US",
  "timeRange": "today 3-m",
  "summary": "93 points, 8 keywords from 2 comparisons, averages chatgpt 67, gemini 27, claude 18, perplexity 1, copilot 5, deepseek 2, grok 4, mistral 0",
  "batchCount": 2,
  "anchorKeywords": ["chatgpt"],
  "batchExploreUrls": [
    "https://trends.google.com/trends/explore?q=chatgpt%2Cgemini%2Cclaude%2Cperplexity%2Ccopilot&date=today+3-m&geo=US",
    "https://trends.google.com/trends/explore?q=chatgpt%2Cdeepseek%2Cgrok%2Cmistral&date=today+3-m&geo=US"
  ],
  "averages": {"chatgpt": 67, "gemini": 27, "claude": 18, "perplexity": 1, "copilot": 5, "deepseek": 2, "grok": 4, "mistral": 0},
  "timeline": [
    {"date": "2026-07-17T00:00:00Z", "formattedTime": "Jul 17, 2026", "values": {"chatgpt": 63, "gemini": 23, "claude": 21, "perplexity": 1, "copilot": 5, "deepseek": 2, "grok": 3, "mistral": 0}, "isPartial": false},
    {"date": "2026-07-18T00:00:00Z", "formattedTime": "Jul 18, 2026", "values": {"chatgpt": 48, "gemini": 19, "claude": 15, "perplexity": 1, "copilot": 2, "deepseek": 1, "grok": 4, "mistral": 0}, "isPartial": false},
    {"date": "2026-07-19T00:00:00Z", "formattedTime": "Jul 19, 2026", "values": {"chatgpt": 45, "gemini": 15, "claude": 12, "perplexity": 1, "copilot": 3, "deepseek": 2, "grok": 4, "mistral": 1}, "isPartial": false}
  ]
}

US metro areas for coffee, tea with regionResolution: "metro":

{
  "type": "interestByRegion",
  "status": "ok",
  "keywords": ["coffee", "tea"],
  "geo": "US",
  "summary": "210 regions, top Cheyenne WY-Scottsbluff NE",
  "resolution": "DMA",
  "valueType": "share",
  "regions": [
    {"geoCode": "759", "geoName": "Cheyenne WY-Scottsbluff NE", "values": {"coffee": 73, "tea": 27}},
    {"geoCode": "678", "geoName": "Wichita-Hutchinson KS", "values": {"coffee": 68, "tea": 32}},
    {"geoCode": "744", "geoName": "Honolulu HI", "values": {"coffee": 70, "tea": 30}}
  ]
}

Cities have no code, so they come with coordinates:

{"geoCode": null, "geoName": "Bagdad", "values": {"hurricane": 100}, "coordinates": {"lat": 30.5988048, "lng": -87.0310767}}

A pasted explore link keeps its own settings, here the Finance category (cat=7):

{
  "type": "interestOverTime",
  "status": "ok",
  "comparisonIndex": 32,
  "keywords": ["mortgage rates", "inflation"],
  "geo": "US",
  "timeRange": "today 12-m",
  "category": 7,
  "inputUrl": "https://trends.google.com/trends/explore?cat=7&date=today%2012-m&geo=US&q=mortgage%20rates,inflation",
  "summary": "53 points, averages mortgage rates 54, inflation 10"
}

Research on pickleball with researchMaxRelated: 5, one related query and the free summary:

{
  "type": "relatedTrend",
  "status": "ok",
  "researchKeyword": "pickleball",
  "query": "best pickleball paddles 2026",
  "relatedType": "rising",
  "relatedRank": 1,
  "relatedValue": 2800,
  "relatedFormattedValue": "+2,800%",
  "isBreakout": false,
  "summary": "53 points, average 40, growth -52.7%",
  "average": 40,
  "growthPercent": -52.7,
  "recentAverage": 34.3,
  "previousAverage": 72.5,
  "growthWindowPoints": 13,
  "timeline": [
    {"date": "2026-09-27T00:00:00Z", "formattedTime": "Sep 27 – Oct 3, 2026", "values": {"best pickleball paddles 2026": 19}, "isPartial": false},
    {"date": "2026-10-04T00:00:00Z", "formattedTime": "Oct 4 – 10, 2026", "values": {"best pickleball paddles 2026": 17}, "isPartial": false},
    {"date": "2026-10-11T00:00:00Z", "formattedTime": "Oct 11 – 17, 2026", "values": {"best pickleball paddles 2026": 15}, "isPartial": true}
  ]
}

{
  "type": "researchSummary",
  "status": "ok",
  "researchKeyword": "pickleball",
  "summary": "5 of 5 related queries with data, highest growth rpm pickleball +19.9%",
  "seedGrowthPercent": -11.1,
  "relatedTrends": [
    {"query": "rpm pickleball", "relatedType": "rising", "status": "ok", "average": 47, "growthPercent": 19.9, "recentAverage": 68.5, "previousAverage": 57.2},
    {"query": "rpm pickleball paddle", "relatedType": "rising", "status": "ok", "average": 44, "growthPercent": 4.8, "recentAverage": 61, "previousAverage": 58.2},
    {"query": "tesla pickleball paddle", "relatedType": "rising", "status": "ok", "average": 18, "growthPercent": -19.6, "recentAverage": 15.5, "previousAverage": 19.2}
  ]
}

How to read the numbers

Google Trends values are relative, not search counts. In interestOverTime, 100 is the peak for the busiest keyword in the comparison, so keywords in one comparison share one scale. In interestByRegion, a single keyword gets a 0 to 100 index per region (valueType: "index"). With 2 or more keywords, each region shows how the searches split between them (valueType: "share"). Rising related queries show percent growth, and Breakout means growth above 5,000%. searchVolume in Trending Now is Google's rounded bucket, such as 50,000 or 500,000.

Pricing

Prices are per 1,000 comparisons or searches.

Event	No discount	Bronze	Silver	Gold
trend_query, a comparison of 1 to 5 keywords that returns data, with its timeline, regions and related queries	$1	$1	$1	$1
trending_search, one Trending Now search returned	$0.25	$0.225	$0.20	$0.175
No data, invalid input or failed fetch	Free	Free	Free	Free

Bronze, Silver and Gold are Apify Store discount tiers, and Apify's Pricing tab shows them as Starter, Scale and Business. Apify decides which tier an account is on. Keyword comparisons cost the same on every tier.

The region and related-query rows come with the comparison at no extra cost, and so does a city or metro breakdown. A pasted explore link is one comparison. With more than 5 keywords, each Google comparison the scraper sends is one trend_query, charged only when it returns data. That is 2 for 6 to 9 keywords, 3 for 10 to 13, and 1 more for every 4 keywords after that, up to 13 for 50. In research, the seed keyword is one trend_query, and so is each related query whose timeline returns data, so researchMaxRelated: 10 costs at most 11 per seed. The researchSummary row is free. Trending Now returns the top 50 searches per country by default, which costs $0.0125 per country, or $0.00875 on Gold. Lower trendingMaxItems to pay for fewer. Set a maximum cost per run in the run options, and the scraper stops when it is reached.

Use it from AI agents

The scraper works as a tool through the Apify MCP server in Claude, Cursor, VS Code and other MCP clients. An agent can ask "is interest in X growing, and where?" and get the timeline, regions and rising searches in one call. With researchKeywords, "what is rising around X?" also returns each rising search's own curve and growth in the same call.

You can also call it through the Apify API or the Python and JavaScript clients, and schedule it or connect it to Make, Zapier and n8n.

More from heyitsgautham

Other actors from the same account.

• Etsy Products & Listings Scraper (Search, Shop, Category) for Etsy listings with price, sale price, shop and rating, for product and price research.
• YouTube Search Results & Rank Tracker Scraper for YouTube search results with rank, to see which videos rank for a keyword that Trends shows rising.
• Trustpilot Reviews & TrustScore Scraper for Trustpilot reviews and each company's TrustScore, for any business on any Trustpilot country site.
• YouTube Video & Shorts Comments Scraper with Replies for the comments and replies on a video or Short, for sentiment analysis and audience research.

FAQ

Why did a comparison come back as no_data? Google Trends only shows keywords with enough search volume. Very rare keywords, small regions and short time ranges often have none. Those rows are free.

Why did a made-up keyword return a few points? For short time ranges and rare keywords, Google sometimes shows a sparse timeline with one or two non-zero points. The website shows the same chart. The scraper returns what Google returns, so that comparison counts as data.

Why did a comparison come back as error? Google Trends sometimes rate limits its data, mostly when many requests reach it at the same moment. The scraper then retries on a fresh residential IP, waiting longer each time, for about two minutes. If Google still refuses, you get an error row that says so, the rest of the run carries on, and that comparison is free. Running it again a few minutes later usually works.

Does it return related topics? No. Google stopped serving related topics to automated clients in 2026 and now returns an empty list, so this scraper returns related queries only.

Why are my numbers slightly different from the website? Google samples its data, so the same query can shift by a few points between requests. That happens on trends.google.com too.

Why does a keyword at 0 or 1 look less exact in a comparison of more than 5? Google rounds every value to a whole number. A keyword far below the top one rounds to 0 or 1, the same as it would on the website next to that top keyword. In the measured pairs, the ratio between two such keywords was off by up to 23%, while the rest were within 3.6%.

Which time zone are the dates in? UTC.

Does it use an official Google API? No. It reads the same public data the Google Trends website loads, without logging in.

Limitations

• Up to 5 keywords per comparison, the same limit as the website, or up to 50 with moreThanFiveKeywords.
• Related topics are not available (see the FAQ).
• Cities appear only where Google has enough data, and metro areas exist only for the US.
• A Trends link that gives each keyword its own location or date returns a free invalid_input row.
• Trending Now covers the countries Google lists on trends.google.com/trending.