What it does

Enter a Wikimedia Commons category. The tool reads which pages across Wikimedia projects use the files in it, how many people viewed those pages each month, and how that usage is spread across languages, projects, and file types — all live, queried directly from Wikimedia's public APIs. Nothing is stored between runs.

Using it

• Category accepts a bare name, Category:Name, or a full commons.wikimedia.org URL.
• Subcategory depth controls how far it descends into subcategories to find files. Higher depth is slower and, for categories with very large subcategories, can take a while.
• Meta/coordination wikis (meta, outreach, wikitech, incubator) are excluded by default since they aren't reader-facing content sites. Toggling this re-filters instantly without a new query.
• Copy link / QR code share the exact report (category, depth, dates, filter) as a short link. If the short-link service isn't reachable — e.g. you're viewing this file directly rather than through the deployed tool — it falls back to the full URL automatically.
• Every chart has SVG / PNG / CSV export buttons, and there's a full wikicode export for pasting the whole report on-wiki.

Reading the numbers

• Pageviews in period counts real readers of pages carrying the category's media — a reach number, not an isolated attribution test of the media by itself.
• Wikimedia's pageview processing lags by several weeks. If the last point on the monthly chart drops sharply below the trend, that month isn't final yet — move "To" back a month and re-run.
• If you see a warning under the Pageviews tile ("N of M page(s) failed to fetch..."), some lookups failed after retries and were counted as zero. The total shown is a floor, not exact — re-run to retry.
• Documenters counts unique uploaders across the files in the category (from each file's latest revision).

Changelog