Topic Trends - a Google Trends Alternative Built on Wikipedia Data

See how much interest any topic gets over time and compare topics on one 0-100 scale, the way Google Trends does - but from an open, official data source that never blocks you: Wikipedia pageviews. Wikipedia is read billions of times every month, and Wikimedia publishes how often each article was read, per day, since July 2015.

For every keyword you get the interest-over-time timeline (daily, weekly or monthly), a 0-100 value normalised across all keywords of the run, total and average views, the peak, a rising / stable / falling trend, spike days, and interest by language (how the topic is read in German, Japanese, Spanish... Wikipedia). A second mode lists what is trending today in any country or language.

No Google scraping, no captchas, no proxies, no "429 Too Many Requests" from Google - and $3 per 1,000 topics.

What you can use it for

• Trend research and content planning - which of your topic ideas is growing, and when it peaks each year (seasonality back to 2015).
• Brand and competitor tracking - compare brands, products, people or shows on one scale, week by week.
• Market and investment research - attention to companies, technologies, drugs or crypto assets over time, and the day a story broke (spikes).
• News and social monitoring - the daily list of what people in the US, UK, India, Germany or Japan are reading about.
• Data journalism and academic work - a documented, reproducible public dataset (CC0), unlike Google Trends' re-sampled index.
• AI agents - one tool call through the Apify MCP server answers "is interest in X growing?" with numbers.

Why not just scrape Google Trends?

Google Trends has no official API. Scrapers break whenever Google changes the page, get rate-limited, and return a sampled index that changes between two identical requests. This Actor uses the official, free Wikimedia Pageviews API and the MediaWiki API, so results are stable, repeatable and allowed to be reused (CC0).

Be clear about the difference: Google Trends counts searches, this Actor counts people reading about a topic. The two move together for most topics (news events, product launches, seasonal interest), but they are not the same number, and there is no per-city breakdown - the closest open equivalent to "interest by region" is interest by language edition.

Input

Field	What it does	Default
mode	topics (interest over time) or trending (top topics of a day)	topics
keywords	One per line: a keyword (bitcoin), an exact article title (Apple Inc.) or a Wikipedia URL. Up to 200.	ChatGPT, Bitcoin, Taylor Swift
language	Wikipedia edition used to match keywords: en, de, fr, es, ja...	en
timeRange	Past 7 / 30 / 90 days, past 12 months, past 5 years, 2015-present, or custom	past 90 days
startDate / endDate	Custom period (YYYY-MM-DD)	-
granularity	auto (daily up to 3 months, weekly up to ~2 years, then monthly), daily, weekly, monthly	auto
includeRedirects	Also count old names and alternative titles (so a renamed article does not "start from zero")	true
includeLanguages / maxLanguages	Interest by language edition, largest editions first	true / 10
platform	All devices, desktop, mobile web or mobile app	all
trendingCountries	Trending mode: country codes, e.g. US, GB, IN, DE, JP	-
trendingProjects	Trending mode: language editions, e.g. en, de	the language edition
trendingDate / maxTrending	Trending mode: day (default = latest published) and list length	latest / 25

Tip: a keyword is matched to one article: first the exact title, otherwise the top search hit (disambiguation pages are skipped). apple gives the fruit - write Apple Inc. for the company. The matched article is in every result (article, articleUrl, matchedBy), so you can check it.

{
    "keywords": ["Nike, Inc.", "Adidas", "Puma (brand)", "New Balance", "Under Armour", "Lululemon", "Skechers", "Asics", "Hoka One One", "On (company)"],
    "timeRange": "past12Months"
}

Output

Real results of the input above (cloud run, 2026-09-23, 12 months, weekly):

Keyword	Matched article	Avg interest (0-100)	Views (12 mo)	Trend	Peak week	Top languages
Nike, Inc.	Nike, Inc.	50.3	904,814	stable (-8.6%)	2026-09-06	en 56.8%, es 7%, de 6.7%
Adidas	Adidas	47.9	860,495	stable (+5.9%)	2026-06-14	en 45.3%, de 12%, es 9.7%
Lululemon	Lululemon	34.3	617,387	falling (-30.5%)	2026-02-01	en 84.4%, ja 5.6%, de 4.1%
On (company)	On (company)	25.3	455,592	stable (+16.1%)	2026-09-13	en 65.1%, fr 15.4%, de 7.3%
Puma (brand)	Puma (brand)	24.0	433,533	stable (-0.7%)	2026-06-28	en 47.5%, de 11.9%, es 8%
Hoka One One	Hoka	20.4	368,104	stable (-17.1%)	2025-10-19	en 68.1%, fr 9%, es 5%
Asics	Asics	17.9	323,532	stable (-3.1%)	2025-09-28	en 47.8%, ja 19.8%, de 6.3%
New Balance	New Balance	16.3	292,884	stable (-11.9%)	2025-09-28	en 54.6%, ja 8.6%, de 6.9%
Under Armour	Under Armour	14.1	253,253	falling (-28.2%)	2025-12-28	en 55.5%, fr 9%, de 7%
Skechers	Skechers	11.5	206,731	stable (-4.1%)	2025-09-28	en 55.2%, de 8.6%, es 8.6%

One full item (default input, 90 days daily; the timeline has one entry per day - shortened here):

{
    "type": "topic",
    "keyword": "Bitcoin",
    "article": "Bitcoin",
    "articleUrl": "https://en.wikipedia.org/wiki/Bitcoin",
    "language": "en",
    "description": "Decentralized digital cryptocurrency",
    "wikidataId": "Q131723",
    "matchedBy": "title",
    "startDate": "2026-06-24",
    "endDate": "2026-09-21",
    "granularity": "daily",
    "totalViews": 237187,
    "averageViewsPerDay": 2635.4,
    "interestScore": 1,
    "peakDate": "2026-08-28",
    "peakViews": 5172,
    "latestValue": 1,
    "changePercent": 10.5,
    "trend": "stable",
    "spikes": [],
    "timeline": [
        { "date": "2026-06-24", "views": 2550, "value": 1 },
        { "date": "2026-06-25", "views": 3140, "value": 1 },
        { "date": "2026-09-20", "views": 2517, "value": 1 },
        { "date": "2026-09-21", "views": 2483, "value": 1 }
    ],
    "interestByLanguage": [
        { "language": "en", "article": "Bitcoin", "views": 237187, "sharePercent": 57.7 },
        { "language": "de", "article": "Bitcoin", "views": 39845, "sharePercent": 9.7 },
        { "language": "ru", "article": "Биткоин", "views": 31731, "sharePercent": 7.7 },
        { "language": "zh", "article": "比特币", "views": 23309, "sharePercent": 5.7 },
        { "language": "ja", "article": "ビットコイン", "views": 12726, "sharePercent": 3.1 }
    ],
    "redirectViewsIncluded": 1045,
    "platform": "all-access",
    "source": "Wikimedia Pageviews API (human traffic only), CC0",
    "error": null
}

value and interestScore are relative, exactly like Google Trends: 100 is the highest point of any keyword in the run (here ChatGPT, ~70,000 views a day), so Bitcoin sits at 1. views are the absolute numbers, so you can always re-scale. isPartial: true marks a week or month at the edge of the period.

Trending mode ({"mode": "trending", "trendingCountries": ["US", "GB", "IN"], "maxTrending": 10}), real rows for 2026-09-21:

List	#	Topic	Views (rounded up)	Description
country:US	1	Lizzie Borden	433,900	American woman acquitted of murder (1860-1927)
country:US	2	Cindy Crawford	255,500	American model and actress (born 1966)
country:GB	3	The Blame (TV series)	49,800	British television series
country:IN	2	India at the 2026 Asian Games	97,700	Sporting event delegation
country:IN	3	2026 Asian Games	84,400	Multi-sport event in Aichi and Nagoya, Japan

Main pages, search pages and other non-article pages are removed. Per-country views are rounded up to the nearest 100 by Wikimedia (privacy); per-edition lists have exact views.

Pricing

Pay per event, no subscription:

Event	Price
Topic analysed (one keyword with its full timeline, stats and language breakdown)	$0.003
Trending article (one row of a trending list)	$0.0005
Actor start	$0.00005

You run	You pay
10 brands, 12 months	$0.03
100 keywords, 5 years	$0.30
Daily top 25 for 5 countries, for a month	25 x 5 x 30 x $0.0005 = $1.88

Keywords that match no article and failed lookups are returned with an error and not charged. Set "Maximum cost per run" and the Actor stops cleanly when it is reached.

FAQ

How far back does the data go? Daily data from 2015-07-01 to yesterday (UTC). Today's numbers are published by Wikimedia the next day.

Are bots counted? No. Only human traffic (agent=user) is counted; Wikimedia filters out known crawlers and automated traffic.

Why do my numbers differ from Google Trends? Different signal: searches on Google vs. reading on Wikipedia. Wikipedia interest also leans towards topics people want explained (people, events, companies, science) rather than navigational searches ("facebook login").

How fast is it? About 5-10 seconds for 3 keywords, about a minute for 10 keywords over 12 months with the language breakdown. Wikimedia limits how fast any client may ask, and the Actor paces itself so that runs finish instead of failing.

Can I schedule it? Yes - use Apify Schedules (for example, daily trending lists for your countries) and send the dataset to Google Sheets, Slack or a webhook with Apify integrations.

Data source and license

Pageview data: Wikimedia Foundation, Pageviews API, released under CC0. Article titles and descriptions: Wikipedia / Wikidata. This Actor is not affiliated with Google or the Wikimedia Foundation.

Related Actors by Kanto Labs

• Website Traffic Rank Checker - how popular a website is, worldwide and in 40 countries
• Domain Authority Checker - open-data authority score for any domain
• Company Enrichment - Logo, Socials, Industry & Size
• Free Image Search - Royalty-Free Photos - pictures for the topics you write about