Verify the exit you were sold, against sources you can name.

Does the exit address support the description I bought?

Check the actual exit rather than the gateway, compare source-reported role and location with the seller’s words, and export the dated differences. An inconsistency supports a specific question, not a certificate of fraud or suitability.

Source availability depends on the current deployment and each returned record. These review dates do not establish when a particular IP was observed. Method axis-scope-2026-09-15 · Score ipjury-score-v3.

You paid for an exit described as residential. Compare that description with the named network and location evidence, recording the exact claim you are checking.

Check the exit, not the endpoint

The address you connect to is often a gateway. The one that matters is the address traffic leaves from — the exit the destination sees. Route a request through the proxy, check the address it reports, then read the record in this order.

1. Whose ASN announces it? A consumer ISP or mobile carrier announcing the address is consistent with a residential or mobile sale. A hosting label is a reason to ask for clarification. Neither label alone certifies the exit path or exclusive use.
2. Does the registration country match the geolocated country? Registry allocation and usage estimate are two separate claims, and IPJury shows them separately. A mismatch is a question to put to the seller, not proof of fraud — allocations legitimately cross borders.
3. What does the PTR hostname suggest? Operator-controlled naming carrying pool, dsl, cable, or mobile language points one way; naming carrying vps, server, cloud, or a hosting brand points another. PTR is a weak clue and IPJury labels it as one. It corroborates the ASN reading; it does not override it.
4. Read role, anonymity, and abuse as three answers. A residential role does not mean no anonymity evidence. An anonymity finding does not mean an abuse record exists. An abuse record does not tell you the role. Collapsing the three into "is it good" is exactly how a seller's claim survives a check it should have failed.
5. Read coverage and abstentions before you conclude anything. If half the jurors abstained, a quiet record is a thin record. "No evidence in covered sources" is not "clean" — it means the sources that answered returned no finding.

When sources disagree, that is the finding

Different network descriptions can reflect different scopes or collection methods; the report names its own returned claims. A network can lease space to a hosting tenant while remaining a consumer ISP on paper. A residential-proxy operator can sit inside genuine subscriber space. One dataset weights the allocation record, another weights observed service behavior, and they arrive at different labels from the same address.

IPJury does not average that away. It shows which juror said what, and on what basis. For a purchase decision, split evidence beats false unanimity: an exit described as hosting by a covered source gives you a specific question, but does not establish how a destination will classify it.

Turn the check into an artifact

A screenshot of a number proves nothing to a seller, and they know it. A record that names its sources, states each source's basis, and can be re-run by the other side is a different kind of message. Export the verdict as text, Markdown, or BBCode for a ticket or forum thread, or as a PNG evidence card for a chat window. Include the address and checked time so the seller can investigate. A new lookup may return newer evidence and is not guaranteed to reproduce the earlier result.

Two caveats worth stating when you send it. IP data moves, so a check is evidence about a moment rather than a permanent property. And no record obliges a seller or a platform to agree with it.

Why a neutral record carries weight

Evaluate the record by its sources, dates, scope and exclusions, rather than by who gives the most reassuring number. This lookup does not purchase a proxy or certify the seller’s claims. It reports what named sources returned — including when they contradict each other, and when they say nothing at all.