Trends · Apify actor

Interest over time for any topic, from Wikipedia pageviews

Compare topics on one 0-100 scale, the way Google Trends does, from an open and official source: Wikipedia pageviews, published daily since July 2015. Each keyword gets its timeline, peak, rising or falling trend, spike days and interest by language edition. A second mode lists what people in any country are reading about today.

$0.003 per topic$0.0005 per trending articleFailed inputs are freeLaunching soon

Launching soon. Topic Trends is built, tested and priced, and is waiting to go public on the Apify Store. The button below opens its Store page, which will be live shortly.

Run it on Apify See pricing

What it does

Topic Trends, in short

Daily, weekly or monthly timeline for each keyword, normalised 0-100 across all keywords of the run, with absolute views
Total and average views, peak, latest value, change percent and a rising / stable / falling trend
Spike days, and interest by language edition (the closest open equivalent to interest by region)
Redirects counted, so a renamed article does not look like it started from zero
Trending mode: the most-read articles of a day per country or language edition
Official Wikimedia APIs only; the pageview data is CC0

Who it's for

Built for

Content planningWhich topic ideas are growing, and when they peak each year, back to 2015.

Brand and competitor trackingCompare brands, products, people or shows on one scale, week by week.

Market researchAttention to companies, technologies or assets over time, and the day a story broke.

News monitoringThe daily list of what readers in the US, UK, India, Germany or Japan are reading about.

Example

Input and output

One dataset row per result, downloadable as JSON, CSV or Excel, or read through the Apify API.

Input

{
  "keywords": ["Nike, Inc.", "Adidas", "Puma (brand)", "New Balance"],
  "timeRange": "past12Months"
}

From a real cloud run on the default input, shortened. value is relative: 100 is the highest point of any keyword in the run (there, ChatGPT), which is why Bitcoin sits at 1. views are absolute.

Output (one row)

{
  "type": "topic",
  "keyword": "Bitcoin",
  "article": "Bitcoin",
  "matchedBy": "title",
  "startDate": "2026-06-24",
  "endDate": "2026-09-21",
  "granularity": "daily",
  "totalViews": 237187,
  "averageViewsPerDay": 2635.4,
  "peakDate": "2026-08-28",
  "changePercent": 10.5,
  "trend": "stable",
  "timeline": [
    { "date": "2026-06-24", "views": 2550, "value": 1 }
  ],
  "interestByLanguage": [
    { "language": "de", "article": "Bitcoin", "views": 39845, "sharePercent": 9.7 }
  ],
  "error": null
}

Pricing

Pay per result. Failures are free.

$0.003topic analyzed · $3.00 per 1,000

No subscription, no monthly minimum. Apify platform compute is included in the price; Apify also adds a tiny start fee per run ($0.00005 per GB of run memory).

Example costs
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	$1.875

Worked example. You compare 40 product names over 5 years and 2 of them match no Wikipedia article.38 topics × $0.003 = $0.114
2 unmatched = $0.00

Not charged:

Keywords that match no article
Failed lookups and trending lists with no data yet

Trending mode is charged per row returned: $0.0005 per trending article. You can set Maximum cost per run in the run options and the actor stops cleanly when it is reached.

FAQ

Questions

Why do the numbers differ from Google Trends?

Different signal: Google Trends counts searches, this counts people reading about a topic on Wikipedia. The two move together for most news, product and seasonal topics, but they are not the same number.

How far back does the data go?

Daily data from 2015-07-01 to yesterday (UTC). Wikimedia publishes each day's numbers the next day.

How is a keyword matched?

To one article: the exact title first, otherwise the top search hit, skipping disambiguation pages. apple gives the fruit; write Apple Inc. for the company. The matched article is in every result.

Are bots counted?

No. Only human traffic is counted; Wikimedia filters out known crawlers and automated traffic.

Is there interest by city or region?

No. The closest open equivalent is interest by language edition, which is included.