Google Trends Scraper & API: keyword trends, related queries and trending searches

This Google Trends scraper works as an unofficial Google Trends API: scrape Google Trends for any keyword and get interest over time, interest by region, related queries, related topics and today's trending searches as clean JSON, CSV or Excel. Track keyword trends, compare up to 5 terms on one scale, and spot rising and breakout searches with no API key and no browser.

For every keyword, or group of up to 5 keywords compared against each other, you get:

• Interest over time: the 0-100 timeline Google shows, with dates, plus the average, peak, latest value and % change for each keyword
• Interest by region: 0-100 by country, state/province, US metro (DMA) or city
• Related queries: top and rising, including "Breakout" terms
• Related topics: top and rising, with Knowledge Graph IDs, when Google has them
• Trending now: Google's current trending searches for any country, with approximate search volume and news links

It's built to run without failing. Everything goes over plain HTTP, with no headless browser. Each session gets a Google cookie first. When Google rate-limits a request (HTTP 429), the scraper rotates to a new session and proxy IP and backs off with jitter. A failed search is retried on its own, so it never takes down the rest of the run.

How to use the Google Trends scraper

1. Enter your keywords in Search terms (one per line, or comma-separated on one line to compare them), and optionally a country, time range, or countries for Trending now.
2. Click Start. A handful of searches usually finishes in well under a minute.
3. Download the results as JSON, CSV or Excel, or call the actor through the Apify API and get the dataset in your own code.

Why this one

Google Trends is notoriously hard to scrape: it answers 429 Too Many Requests to anything that looks automated. Most Trends scrapers fail a large share of their runs because of it. This actor was written around that problem:

• One small HTTP flow per search: cookie warm-up, explore, then the data widgets
• A new session and IP after a 429, plus short in-session retries for widget calls that hit a transient limit
• Results are saved per search as soon as they're ready, so a slow or failing search never blocks the others
• A RUN_SUMMARY record lists what came back for each search and what failed and why

Use cases

• SEO and content planning: find rising and breakout queries before your competitors do
• Product and e-commerce research: seasonality, demand by state, comparing brands (e.g. iphone, samsung galaxy)
• Market and investment research: attention indicators for tickers, crypto and brands
• Newsrooms and social media teams: daily trending searches with the news stories behind them
• Dashboards: schedule the actor and feed the output into Sheets, Looker Studio or your own database

Input

Field	Default	Description
searchTerms	-	One search per line. Put up to 5 comma-separated terms on a line to compare them on one scale: chatgpt, gemini, claude
exploreUrls	-	Links pasted straight from trends.google.com/trends/explore. Keywords, country, dates, category and search type are read from each URL.
geo	Worldwide	US, GB, DE, US-CA, ...
timeRange	Past 12 months	Past hour, 4 hours, day, 7 days, 30 days, 90 days, 12 months, 5 years, or 2004-present
customTimeRange	-	2024-01-01 2024-12-31
category	0 (all)	Google Trends category ID, e.g. 7 Finance, 71 Food & Drink
searchProperty	web	web, images, news, youtube, shopping
includeInterestOverTime / includeInterestByRegion / includeRelatedQueries / includeRelatedTopics	on	Turn off what you don't need to make runs faster
regionResolution	AUTO	COUNTRY, REGION, DMA (US metros) or CITY
trendingNowGeos	-	Country codes to fetch "Trending now" searches for
proxyConfiguration	Apify Proxy	Switch to RESIDENTIAL if you see many rate-limit retries

Input example

{
    "searchTerms": ["bitcoin", "chatgpt, gemini, claude", "air fryer"],
    "geo": "US",
    "timeRange": "today 12-m",
    "trendingNowGeos": ["US", "GB"]
}

Output example

You get one result per search (a line in searchTerms or an explore URL), with every data type nested inside it, plus one result per trending search if you asked for trending now. This one is from a real run: US, past 12 months (arrays shortened).

{
    "type": "search",
    "searchTerm": "chatgpt, gemini, claude",
    "keywords": ["chatgpt", "gemini", "claude"],
    "geo": "US",
    "geoLabel": "US",
    "timeRange": "today 12-m",
    "category": 0,
    "searchProperty": "web",
    "hasData": true,
    "keywordStats": [
        { "keyword": "chatgpt", "average": 77.6, "peakValue": 100, "peakDate": "2025-10-19T00:00:00.000Z", "latestValue": 76, "changePercent": -20.2 },
        { "keyword": "gemini", "average": 26.9, "peakValue": 35, "peakDate": "2026-04-12T00:00:00.000Z", "latestValue": 33, "changePercent": 20.9 },
        { "keyword": "claude", "average": 17.9, "peakValue": 34, "peakDate": "2026-05-24T00:00:00.000Z", "latestValue": 19, "changePercent": 263.6 }
    ],
    "averageInterest": { "chatgpt": 78, "gemini": 27, "claude": 18 },
    "interestOverTime": [
        { "date": "2026-09-13T00:00:00.000Z", "formattedTime": "Sep 13 – 19, 2026", "values": { "chatgpt": 76, "gemini": 33, "claude": 19 }, "isPartial": false },
        { "date": "2026-09-20T00:00:00.000Z", "formattedTime": "Sep 20 – 26, 2026", "values": { "chatgpt": 82, "gemini": 36, "claude": 21 }, "isPartial": true }
    ],
    "interestByRegion": [
        { "geoCode": "US-CA", "geoName": "California", "values": { "chatgpt": 57, "gemini": 26, "claude": 17 } },
        { "geoCode": "US-DC", "geoName": "District of Columbia", "values": { "chatgpt": 58, "gemini": 20, "claude": 22 } }
    ],
    "regionResolution": "REGION",
    "relatedQueriesTop": [
        { "query": "what is chatgpt", "keyword": "chatgpt", "value": 100, "formattedValue": "100", "isBreakout": false, "link": "https://trends.google.com/trends/explore?q=what+is+chatgpt&date=today+12-m&geo=US" }
    ],
    "relatedQueriesRising": [
        { "query": "how to use chatgpt effectively", "keyword": "chatgpt", "value": 2050, "formattedValue": "+2,050%", "isBreakout": false, "link": "https://trends.google.com/trends/explore?q=how+to+use+chatgpt+effectively&date=today+12-m&geo=US" }
    ],
    "relatedTopicsTop": [],
    "relatedTopicsRising": [],
    "exploreUrl": "https://trends.google.com/trends/explore?q=chatgpt%2Cgemini%2Cclaude&geo=US&date=today+12-m&hl=en-US",
    "scrapedAt": "2026-09-24T05:14:08.994Z"
}

A trending-now result:

{
    "type": "trending",
    "geo": "US",
    "rank": 1,
    "title": "dodgers schedule",
    "approxTraffic": "1000+",
    "approxTrafficNumber": 1000,
    "publishedAt": "2026-09-24T04:40:00.000Z",
    "picture": "https://encrypted-tbn2.gstatic.com/images?q=tbn:...",
    "pictureSource": "MLB.com",
    "news": [
        { "title": "The teams no one wants to play in October", "url": "https://www.mlb.com/news/scariest-mlb-teams-to-play-in-october-2026", "source": "MLB.com", "picture": "https://encrypted-tbn2.gstatic.com/images?q=tbn:..." }
    ],
    "scrapedAt": "2026-09-24T05:14:09.089Z"
}

keywordStats.changePercent compares the average of the last quarter of the period with the average of the first quarter. Partial (still-running) periods are left out of the stats.

The dataset has three views: Overview, Interest over time and Trending now. The run's key-value store also holds a RUN_SUMMARY with per-search counts, retries and any failures.

Pricing

Pay per result: $2.50 per 1,000 results. One result = one search, or one trending search. A comparison of up to 5 keywords on one line counts as one result, and it includes the timeline, regions, related queries and related topics. Searches that fail after all retries are not charged.

• 100 results = $0.25
• 1,000 results = $2.50
• 10,000 results = $25

The Apify free plan includes $5 of monthly credit, enough for about 2,000 results.

Integrations

• Make, Zapier and n8n: start runs and pass the Google Trends data into any workflow
• Google Sheets: export the dataset straight into a spreadsheet, or refresh it on a schedule
• Apify API: REST API plus official JavaScript and Python clients; run the actor and fetch results from your own code
• Webhooks: get notified when a run finishes and pull the results automatically
• Schedules: run daily or hourly to track keyword trends over time
• MCP for AI agents: through the Apify MCP server (https://mcp.apify.com), Claude, ChatGPT, Cursor and other agents can call this actor as a tool and read Google Trends data directly

FAQ

Is it legal to scrape Google Trends? The actor only collects publicly available, aggregated data that anyone can see on trends.google.com, and it contains no personal data. You are still responsible for how you use the results: follow Google's terms of service and the laws that apply to you (such as GDPR). This is not legal advice; ask a lawyer if you're unsure.

How does it avoid blocks? It uses Apify Proxy by default and a session pool: each session first gets its own Google cookie. When Google answers with a rate limit (HTTP 429/403), that session is retired and the search is retried on a fresh session and proxy IP with exponential backoff and jitter (up to 8 retries by default, adjustable to 20). Widget calls that hit a short, transient limit get two quick in-session retries first. Concurrency defaults to 3 parallel searches to stay under Google's limits.

What are the limits? Up to 5 keywords per comparison (Google's own limit; extra terms on a line are dropped). Values are Google's relative 0-100 index, not absolute search volumes. Trending now returns what Google's trending feed lists for each country, typically about 10 searches per country. Related topics and related queries only appear when Google has enough volume. Concurrency is capped at 10.

Why are the numbers 0-100 rather than search volumes? That's how Google Trends reports interest. 100 is the peak for the chosen terms, place and time. Values are only comparable within a single search, so put terms on one line (comma-separated) to compare them on the same scale.

Why do I get slightly different numbers from the website? Google Trends works from a sample of searches. Numbers can move by a few points between requests, even in the browser.

Some searches have empty related queries or topics. Is that a bug? No. Google only shows related queries and topics when there is enough search volume. Related topics have become rare in Google's own responses lately. When Google returns none, you get an empty array rather than an error. hasData: false means Google had too little volume for the term.

I see "rate-limited ... rotating session" warnings in the log. That's normal and handled. Google Trends throttles aggressively, and the actor retries on a fresh IP. If a search still fails after all its retries, it appears in RUN_SUMMARY.failures and no result is charged for it. Many failures usually mean the proxy pool is exhausted. Switch the proxy to the RESIDENTIAL group, or lower maxConcurrency.

Can I get city- or metro-level data? Yes. Set geo to a country and regionResolution to CITY, or to DMA for US metros.

Can I track a topic instead of a search term? Yes. Put a Knowledge Graph ID like /m/0vpj4_b in searchTerms, or paste an explore URL that uses a topic.

How fast is it? In our tests, 9 searches plus 2 trending countries (29 results) took about 20 seconds.