Domain-status fields: API v4

A category alone does not tell you whether a domain is a working site. The opt-in domain-status block does.

Add include_status=true:

curl -X POST "https://website-categorization.whoisxmlapi.com/api/v4/categorize?apiKey=YOUR_API_KEY&include_status=true" \
     -H "Content-Type: application/json" \
     -d '{"domain": "example.com"}'

The response gains seven keys. Below is how a parked domain returns them. They are absent from the default response, not null — check for key presence, not for a null value.

{
  "domain_status": "parked",
  "is_parked": true,
  "staleness": null,
  "last_known_category": null,
  "classified_at": "2026-09-11T09:15:00+00:00",
  "v3_created_date": null,
  "v3_registration_age_days": null
}

The same parameter works on the bulk endpoint, where it appends the block to every result row.

The seven fields

Field Type Description
domain_status string / null What the domain is. See the table below.
is_parked boolean true when the status implies parking.
staleness string / null Age tier of last_known_category. Currently always null — see below.
last_known_category object / null Reserved. Currently always null — see below.
classified_at string / null When the verdict you are reading was produced, ISO-8601 UTC.
v3_created_date string / null Always null — kept for compatibility; do not build on it. See below.
v3_registration_age_days integer / null Always null — kept for compatibility; do not build on it. See below.

domain_status

Value Meaning
active A live classification — model, curated or rule path.
parked A parking page rather than a site.
inactive Not serving content — the domain name does not exist, or the site is gone or a placeholder. Usually safe to drop; see the known issue below.
expired The registration has lapsed.
newly_registered Registered recently and not yet serving meaningful content.
unreachable We got no answer within our fetch budget — a timeout, connection reset or refusal, TLS or renderer failure. A real site we could not read today; retry later.
blocked The host refused us — anti-bot protection or similar.
error The server answered with a 5xx — a CDN origin timeout (Cloudflare 522/524) included.
suppressed Removed at the domain owner's request; contact support@whoisxmlapi.com.

This list is exhaustive as of this release. A value you do not recognise means you are reading a response from a newer release than this page describes.

blocked and unreachable are not parked. They are real sites we could not read, and is_parked stays false for them. Only parked and inactive set is_parked: true. This distinction matters if you are culling a domain list: treating "we could not read it" as "it is dead" will throw away live sites.

A known issue: an inactive domain can be live at www. When a site's bare domain name does not resolve, we can answer inactive, and so is_parked: true, even though its www. address serves the site. is_parked: true covers inactive as well as parked, so do not cull on it alone. Before you drop an inactive domain, check its www. address yourself: this API reads www.example.com as example.com, so asking it again returns the same answer.

Billing

Parked and inactive results charge 0 credits — with or without include_status. A parked domain is never billed. Neither is any other result with no category — a blocked, unreachable or error result among them (see Sites we cannot read, below).

Requesting the block never changes what a call costs.

classified_at

When the verdict you are reading was produced, not when you asked. A cached result returns the original classification time. Pair it with the cache window described on Making requests — up to 30 days, shorter for event-driven categories — to reason about freshness.

last_known_category and staleness — reserved, currently null

last_known_category is reserved for a domain's most recent prior classification: a label and a date only, never retained page content. staleness is the age tier of that label.

Both are currently null on every response — and not merely rare: the leg that would populate them is switched off in production and the store behind it returns nothing, so a populated value is unreachable today, not unlikely. The key is present when you opt in, and the value is null. Treat the field as reserved: build your parser so it tolerates the object appearing later, and do not build logic that depends on a value today.

When the field is populated it will carry tier1, tier2, classified_at and a confidence de-rated by age, with staleness reporting the tier that de-rating came from. Any change here ships additively, with a changelog entry.

Sites we cannot read — no historical fallback

From the 2 October 2026 update the service no longer serves historical-fallback labels.

A domain_status of blocked, unreachable or error means we tried to read the site and could not: anti-bot protection blocked the fetch, the host did not answer, or the server answered with an error. A result with one of these statuses carries no category: categories is empty, primary_category is null, and the call is not billed. A domain answered without reading the site — from our curated list, a domain-name rule or the cache — keeps its category and bills normally, whether or not the site would block us.

The bulk data feed can still carry a label for such a site from an earlier read. The feed's classified_at dates that label: when we last produced or served it, not when we last read the site.

Until the 2 October 2026 update such a result could instead carry the label the legacy WhoisXML v3 service returned for the domain, marked source.is_historical_fallback: true. What remains of that:

  • source.is_historical_fallback stays in the response and is always false. Code that branches on it keeps working.
  • v3_created_date and v3_registration_age_days are always null. Both stay in the response for compatibility; do not build on them.
  • last_known_category is unchanged — reserved and currently null, as described above.

Worked patterns

Cull dead domains from a list, keeping sites we merely could not read:

status = row.get("domain_status")
if status in {"parked", "expired"}:
    drop(row)          # genuinely not a site
elif status == "inactive":
    check_www(row)     # usually gone; check its www. address yourself (see above)
elif status in {"blocked", "unreachable", "error"}:
    retry_later(row)   # a real site we could not read today
else:
    keep(row)