YepAPI
SERP UPDATEOctober 31, 2026·New API

Google SERP API: real domains for Google's new goto links, plus time filters and newest-first sorting

Google's goto redirect links no longer break rank tracking: every result keeps its real domain. New time, sort_by_date and search_param options for fresh results.

We just shipped two changes to the Google SERP API. Google has started wrapping organic result links in `google.com/goto?url=...` redirects, which made many results look like they came from google.com. `/v1/serp/google` now recovers the real domain for those results, so rank tracking and domain matching keep working. And you can now ask Google for fresh results only: past hour, day, week, month or year, sorted newest first.

What's new

  • Google's goto redirect links: each result's `domain` is now the real site (read from the result's breadcrumb), not google.com
  • `time`: hour, day, week, month or year, the same as Google's Tools → Any time menu
  • `sort_by_date: true`: newest results first
  • `search_param`: pass other Google URL parameters through as-is
  • Search operators like `site:`, `inurl:`, `intitle:` and quoted phrases work in `query`, as they always have

Endpoints in this release

What changed at Google

Google has started sending many organic result links through a `https://www.google.com/goto?url=...` redirect instead of linking straight to the site. The token in that link is encrypted, so the destination can't be read from the URL, and every SERP provider is dealing with it. For anyone tracking rankings, it meant results came back with `google.com` as the domain, so a `site:reddit.com` search or a check for "where does my domain rank" could miss results that were really there.

What we do about it

When a result's link is a Google redirect, we take the real domain from the result's breadcrumb (the green site line under the title), or from its site name. The `domain` field is the real site, so rank tracking, competitor matching and `site:` checks work again. The `url` field stays as the link Google served: it still opens the right page in a browser. A few results that show no site name, mostly some forum results, can still report google.com.

Fresh results: time filters and date sorting

Three new optional fields on `/v1/serp/google`:

  • `time`: `hour`, `day`, `week`, `month` or `year`. Only results from that period.
  • `sort_by_date`: `true` sorts results newest first.
  • `search_param`: extra Google URL parameters, for example `"filter=0"`. Parameters that are ignored upstream (`lr`, `cr`, `as_*`) are rejected with a clear 400 instead of being silently dropped.
# Reddit posts about "seo blog" from the past week, newest first
curl -X POST https://api.yepapi.com/v1/serp/google \
  -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"query":"site:reddit.com \"seo blog\"","time":"week","sort_by_date":true}'

Pricing

No change: the Google SERP API is $0.01 per page of 10 results, the new options cost nothing extra, failed calls are free, and new accounts start with $5 of free credit.

Try the Google SERP API

Run a live search with a time filter and see the real domain on every result.

Open the Google SERP API