Find when a web page changed using the Wayback Machine CDX API (free, no key)

Disclosure: I built the Apify Actor linked at the end and charge per result. The raw endpoints below are free and need no key.

The Internet Archive exposes a CDX server that lists every capture it holds for a URL. Two calls cover most needs.

1. List distinct versions of a page

curl -s "https://web.archive.org/cdx/search/cdx?url=example.com&output=json&collapse=digest&fl=timestamp,statuscode,digest&limit=3"

collapse=digest drops consecutive captures with identical content, so each row is a real change. Checked 2026-10-31: the first rows were timestamps 20020120142510, 20020328012821 and 20030207055228.

Open a capture at https://web.archive.org/web/<timestamp>/<url>; add id_ after the timestamp (/web/<timestamp>id_/<url>) for the raw page without the Wayback banner.

2. Find the capture closest to a date

curl -s "https://archive.org/wayback/available?url=example.com&timestamp=20200101"

Checked 2026-10-31: it returned a 200 capture from 2020-01-01 23:10:47.

Gotchas

• Timestamps are YYYYMMDDhhmmss, in UTC.
• Use from=/to= (years or dates) and filter=statuscode:200 to narrow results.
• collapse=timestamp:6 gives roughly one capture per month.
• Be gentle with request rates; the archive throttles heavy clients.

Hosted version

If you'd rather not script it, I packaged both modes as an Apify Actor (pay per result, no subscription): https://apify.com/northgrid-data/wayback-snapshots-cdx

Feedback welcome in the Actor's issues.