ISPing.me

What Is IP Purity? How to Read the Risk Score

💡 Key takeaways

IP purity refers to how clean an IP address has been in its history of use — never abused, never listed in any blacklist; Risk score is the number that quantifies it: the higher the risk score, the dirtier the IP. Purity = 100 − risk score. It is weighted from multiple signals — proxies, Tor, blacklist hits, datacenter attributes — not the verdict of a single database, so checking the tier once beats trusting any one source.

Keep getting flagged when signing up, or never receiving verification codes? The problem may be your IP's “purity”.

Definition

IP purity = 100 − risk score. The higher the risk score, the more risk signals the IP carries; the higher the purity, the “cleaner” the IP looks — more like an ordinary user in the eyes of platforms' risk-control systems. A high-purity IP is the foundation for going global, warming up social accounts, and avoiding seemingly random bans — but it's not a magic pass: type and ASN are judged separately too.

What makes up the risk score

The risk score combines threat intelligence and network-attribute signals from multiple sources. Common deductions include:

Five risk-score tiers

Lower score, cleaner IP

Pure<20No abuse history — good for sign-ups and linking payment methods
Fairly Pure<35Minor flaws — fine for everyday use
Fair<50Some sharing or light abuse history — be careful with sensitive actions
Suspicious<70Clear risk signals — likely to be flagged by risk controls
High Risk≥70A blacklist regular — switch IPs outright

How clean does each use case need to be?

The thresholds below are commonly used reference values in the industry, not an official standard published by any platform; how the same IP performs on different platforms still depends on each platform's risk controls. The parentheticals are rough mappings to this site's five tiers.

Is your cross-border network setup safe?Run one complete, free in-depth IP quality and risk-score check on ISPing.me →, and say goodbye to seemingly random bans.

What difference does purity make?

What to do when you find your IP has gone bad

If checks show serious blacklist records or a very high risk score for your current IP, don't gamble on it for important tasks — pick the fix that matches your situation:

1. Residential broadband users: restart the optical modem and try to get a fresh dynamic IP

Most residential broadband plans assign dynamic IPs. If your current residential IP is “paying for” a previous user's behavior, the simplest fix is to restart the optical modem or router (usually by unplugging it for a few minutes) and reconnect to request a new dynamic IP. Note this only gives you a chance of getting a new IP — it's not guaranteed: carrier binding, lease times, and CGNAT can all hand you back the same IP. If it doesn't change, try again or contact your ISP. For the trade-offs between dynamic and static IPs, see What's the Difference Between a Static IP and a Dynamic IP?.

2. Professional cross-border users: upgrade your network setup

If your business depends heavily on a stable, high-reputation network, drop cheap, heavily shared proxy nodes and ordinary datacenter VPS plans early, and move to higher-grade network resources: Static residential IP (assigned by a real residential broadband ISP and unchanged for a long time, combining residential weight with the stability of a fixed IP); Dual ISP (registered as an ISP in both datacenter and residential broadband databases, combining datacenter speed with residential resistance to bans). For details, see Why does dual-ISP still show as an “IDC datacenter IP”?, What is a native IP?. Note that both of these cost more — verify the source before buying; don't trust the seller's marketing alone.

3. Self-hosted node operators: appeal to the relevant authorities to get delisted

If you own a dedicated server IP that has unfortunately been blacklisted, you can try a delist request: ① first find out on ISPing.me exactly which database (e.g. AbuseIPDB or Spamhaus) listed the IP; ② make sure the server is clean of malware and has stopped sending abusive traffic; ③ file an appeal on the delist page of that blacklist's official site. Review times vary — 24–48 hours is only a common case, not a promise. For details, see What to do if your IP is blacklisted, and how to check and get delisted.

This site's five risk tiers: what the numbers mean

The risk score has five tiers: under 20 is Pure, under 35 Fairly Pure, under 50 Fair, under 70 Suspicious, and 70+ is High Risk; the lower the score, the more the IP looks like an ordinary user. Note that the risk score is a heuristic estimate (weighted from multiple signals), not an official ban verdict from any platform — how the same IP performs on different platforms still depends on each platform's risk controls. Go back and check the homepage once to see which tier you fall into. Tiers are not set in stone: an IP that is Pure today will go bad if it is used to blast spam, so re-check your tier regularly once you know it. Tiers are based on the risk score alone, not the IP type — residential, datacenter, and business IPs each have their own baseline, so comparing scores across types is meaningless.

Risk scorePurityNotes
Under 20PureNo abuse history — good for sign-ups and linking payment methods
Under 35Fairly PureMinor flaws — fine for everyday use
Under 50FairSome sharing or light abuse history — be careful with sensitive actions
Under 70SuspiciousClear risk signals — likely to be flagged by risk controls
70+High RiskA blacklist regular — switch IPs outright

How blacklist hits are weighted: a qualitative note

Risk scoring looks at hits across multiple public blacklists: a hit in Spamhaus, SORBS, or similar lists is a strong signal that the IP was recently abused; more hits, and fresher hits, mean a higher risk score. But a hit doesn't mean “this IP can never be used again” — blacklists have listing cycles, and after switching to a clean IP block and stopping the abuse, records fade over time. Also, residential and datacenter IPs have different baselines: residential IPs start from a cleaner baseline, so the same hit costs a datacenter IP more — which is why the type classification feeds into the risk score. To see the overall purity distribution of similar IPs, go to the IP Purity Ranking and compare. For how long blacklist records take to fade, see Can an IP on a Blacklist Be Cleaned? How Long Does It Take?.

FAQ

What risk score counts as clean?
By this site's five tiers: under 20 is Pure, under 35 Fairly Pure — both fine for everyday use; above 50, be careful with sensitive actions, and at 70+ we recommend switching IPs outright.

Does low purity guarantee a ban?
Not necessarily. Purity is a heuristic estimate based on public information, and each platform's actual policies differ; low purity just means a higher chance of being flagged by risk controls, not a ban verdict.

How long after switching IPs does it become clean?
It depends on the blacklist's listing cycle and whether the abuse has stopped. After switching to a clean IP block, old records fade over time — but if the new IP itself has a bad history, switching is pointless, so check it first.

Why was my AI sign-up rejected when the risk score isn't high?
AI services weigh ASN and IP type more heavily. Some cloud providers' datacenter IPs can be rejected for their IDC nature even with a low risk score — in that case, switching to a residential or native IP helps more than chasing an even lower score.

Notes

Purity is a heuristic estimate based on public information, not a bank-grade risk score, and each platform's actual policies differ, so treat the result as a reference only. An IP's purity also changes over time — an IP that's clean today will go bad if it's used to blast spam.

Check your IP's purity

Enter an IP on ISPing.me to see the risk score, purity, and a breakdown of every risk signal.

Check now

Related reading:What the IP sharing count meansHow Proxy IPs and Tor Exit Nodes Affect YouWhat to do if your IP is blacklistedWhat cross-border e-commerce requires from an IPPayPal risk control and IPsWhat TikTok requires from an IPWhat ChatGPT requires from an IP

IP Knowledge Base