Score is high in one tool and low in another
Check Check each tool's scale; the bands do not line up and scores cannot be converted.
Fix Compare each score only with that tool's own bands.
What each IP checker measures, why results differ, and how to compare them without trusting one number.
Each checker builds its own database from its own data. Scamalytics says its score comes from fraud feedback reported by a network of operators, and that it applies this feedback to the surrounding subnet, ASN and hosting block, so an address can score higher because of its neighbours. A registry-style flag such as "hosting" comes from who owns the network, not from anyone's fraud reports.
Update times differ too. One database may already have a new owner for an address range while another still shows the old one. Put together, two honest checkers can give different answers for the same address on the same day.
| Item | What it reports | What the source says about it |
|---|---|---|
| Scamalytics | Fraud score 0–100 and four risk bands (0–19, 20–59, 60–89, 90–100) | Built from operator fraud feedback and spread across the subnet, ASN and hosting block; thresholds are a starting point. |
| IPinfo | Yes/no flags: vpn, proxy, tor, relay, hosting | hosting means a hosting, cloud or data center provider; residential-proxy fields sit in a separate endpoint. |
| MaxMind | Anonymous IP flags and connection type | Flags include is_hosting_provider and is_residential_proxy; connection type is Cable/DSL, Cellular, Corporate or Satellite. |
| IP2Location | usage_type code | ISP is a fixed-line ISP, DCH is data center, web hosting or transit, MOB is a mobile ISP. |
| Spur | Classifies VPNs and proxies, including ISP proxies | Describes ISP proxies as residential-looking addresses hosted in data center environments. |
Source: each service's own documentation, read on 11 November 2026 (see the source list). IPQS and other services publish their own scales too; read each on its own site, because scores cannot be converted from one tool to another.
Ask what each tool can actually prove. The country and the organization that owns the ASN are facts you can check in public registries and in several tools; if all of them agree, believe them. A fraud score or a "proxy" flag is an opinion about reputation, and it only matters if the site you care about uses that kind of data.
So the practical rule is: use the type and ASN checks to decide what kind of network the IP is, use the score to spot an address with a bad neighbourhood, and use the real website as the final test. No checker, including ours, can tell you how a platform will treat an IP.
Choose one score tool, one type or ASN tool and one geolocation tool so that you compare different questions, not the same one three times.
Run all lookups within a few minutes and record the date, the IP and each result. Do not mix results from different days.
Write country and ASN owner in one column and score or flag results in another. If the facts disagree, one database is out of date.
A hosting label on an IP whose ASN owner is a consumer ISP is a labelling disagreement. A bad score on an address inside a clean range may be a neighbour effect.
Use the proxy on the real site. That result outranks every checker.
Check Check each tool's scale; the bands do not line up and scores cannot be converted.
Fix Compare each score only with that tool's own bands.
Check Compare ASN owner names and the type codes in the documentation.
Fix Look at the ASN owner first; see the datacenter label guide.
Check Check the country first; city data is less certain.
Fix Read the geolocation guide before you decide the IP is wrong.
There is no free trial. The smallest purchase is one static ISP IP for one day at US$0.50, which is enough to run the five steps above before you lease 30 days at US$7.99. Prices are the same in every country, and the IP does not rotate during the lease.
If the IP does not connect or exits in the wrong country, send Support the order number and your lookups; failed IPs are fixed by reissuing them. A checker disagreement on its own is not a promise of replacement.
None can be called most accurate: they measure different things from different data. Compare the facts they share, such as country and ASN, and treat each score as one opinion.
Each tool has its own scale and data. Scamalytics says its score is built from operator fraud feedback spread across the subnet, ASN and hosting block.
No. Scales and methods differ, so read each score against that tool's own bands.
No. A site uses its own data and rules, so the real site is the final test.
Some services class ISP proxies as proxies even when the address looks residential. We do not promise that any IP reads clean in every checker.
Official documentation · Confirmed on:
Read on 11 November 2026: the score is built from fraud feedback from a network of operators; it is applied across the same subnet, ASN and hosting block; bands 0–19, 20–59, 60–89, 90–100; thresholds are a starting point to tune against your own data.
Official documentation · Confirmed on:
Read on 11 November 2026: boolean fields vpn, proxy, tor, relay and hosting (an IP that belongs to a hosting, cloud or data center provider) plus a service name; residential-proxy fields (is_res_proxy, last_seen, percent_days_seen over a rolling 7-day window) belong to a separate IPinfo Max endpoint.
Official documentation · Confirmed on:
Read on 11 November 2026: flags is_anonymous, is_anonymous_vpn, is_hosting_provider, is_public_proxy, is_tor_exit_node and is_residential_proxy; each is 1 or blank.
Official documentation · Confirmed on:
Read on 11 November 2026: connection_type is one of Cable/DSL, Cellular, Corporate or Satellite.
Official documentation · Confirmed on:
Read on 11 November 2026: usage_type codes include ISP (Fixed Line ISP), MOB (Mobile ISP), DCH (Data Center/Web Hosting/Transit), COM (Commercial), CDN, EDU and others; latitude and longitude default to the capital city when the city is unknown.
Official documentation · Confirmed on:
Read on 11 November 2026: how Spur tells VPNs, residential proxies and ISP proxies apart.
Product observation · Confirmed on:
Product facts file of the site, checked 10 November 2026 against the console, the backend and the price table: per-IP lease prices in US$ (1 PTC = US$1), lease terms, HTTP port 2222 and SOCKS5 port 3333, no free trial, failed assets are reissued after a Support ticket.
Create an account, top up once, then pick rotating residential, static ISP or datacenter IPs and have one running in your browser within minutes.