<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: IP99</title>
    <description>The latest articles on DEV Community by IP99 (@ip99).</description>
    <link>https://dev.to/ip99</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4174889%2Fbd2ba501-747f-48fe-b7f5-e3c0f6d94b3b.webp</url>
      <title>DEV Community: IP99</title>
      <link>https://dev.to/ip99</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/ip99"/>
    <language>en</language>
    <item>
      <title>This Week's Proxy Exits: 10.7M New Addresses, Still Almost All Off Datacenter Space</title>
      <dc:creator>IP99</dc:creator>
      <pubDate>Thu, 19 Nov 2026 13:19:22 +0000</pubDate>
      <link>https://dev.to/ip99/this-weeks-proxy-exits-107m-new-addresses-still-almost-all-off-datacenter-space-5140</link>
      <guid>https://dev.to/ip99/this-weeks-proxy-exits-107m-new-addresses-still-almost-all-off-datacenter-space-5140</guid>
      <description>&lt;p&gt;Datacenter blocklists still look like the default proxy defence in a lot of fraud stacks. Volume came back this week, but the mix did not flip: newly seen proxy exits were still almost all off datacenter space. The datacenter slice ticked up from &lt;strong&gt;2.2%&lt;/strong&gt; in the last posted recap to &lt;strong&gt;2.4%&lt;/strong&gt;, still in the flat ~2% band of recent weeks.&lt;/p&gt;

&lt;p&gt;Between &lt;strong&gt;12–18 November 2026 (UTC)&lt;/strong&gt;, the &lt;a href="https://ip99.com?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP99&lt;/a&gt; risk feed captured &lt;strong&gt;10,705,894&lt;/strong&gt; proxy exit addresses it had not seen in the previous 7 days (&lt;strong&gt;+7%&lt;/strong&gt; vs the previous 7 days). Source page: &lt;a href="https://ip99.com/en/proxy-exits?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;ip99.com/en/proxy-exits&lt;/a&gt; (generated 19 November 2026 UTC). The public page rolls daily over the last 7 complete UTC days; this is today's window, continuing the series after the 9–15 November recap (the two windows overlap on 12–15 November).&lt;/p&gt;

&lt;p&gt;Against that last posted week (&lt;strong&gt;9,914,335&lt;/strong&gt; addresses), volume is up about &lt;strong&gt;+8.0%&lt;/strong&gt;. The page's official &lt;strong&gt;+7%&lt;/strong&gt; compares 12–18 November with 5–11 November — a clean adjacent week, and the same window as the 12 November recap (&lt;strong&gt;10,026,346&lt;/strong&gt;). Much of the step-up versus the last posted recap is compositional: heavier &lt;strong&gt;16–18 November&lt;/strong&gt; days, led by &lt;strong&gt;16 November&lt;/strong&gt; (&lt;strong&gt;1,691,075&lt;/strong&gt;), replaced lighter &lt;strong&gt;9–11 November&lt;/strong&gt; days such as &lt;strong&gt;11 November&lt;/strong&gt; (&lt;strong&gt;1,217,179&lt;/strong&gt;). These figures describe what entered the 7-day mirror, not worldwide proxy supply.&lt;/p&gt;

&lt;h2&gt;
  
  
  What those addresses sit on
&lt;/h2&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Usage type&lt;/th&gt;
&lt;th&gt;Addresses&lt;/th&gt;
&lt;th&gt;Share&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;Residential broadband&lt;/td&gt;
&lt;td&gt;8,425,688&lt;/td&gt;
&lt;td&gt;78.7%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Mobile network&lt;/td&gt;
&lt;td&gt;1,431,350&lt;/td&gt;
&lt;td&gt;13.4%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Gateway / NAT egress&lt;/td&gt;
&lt;td&gt;514,896&lt;/td&gt;
&lt;td&gt;4.8%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Datacenter&lt;/td&gt;
&lt;td&gt;253,933&lt;/td&gt;
&lt;td&gt;2.4%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Education network&lt;/td&gt;
&lt;td&gt;50,157&lt;/td&gt;
&lt;td&gt;0.5%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Unclassified&lt;/td&gt;
&lt;td&gt;29,870&lt;/td&gt;
&lt;td&gt;0.3%&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;About &lt;strong&gt;63%&lt;/strong&gt; were rated high-risk at capture; the rest still carried a proxy-class signal at a lower risk level.&lt;/p&gt;

&lt;p&gt;IP99 counts an address as "captured" when it enters the feed with a proxy-class signal (&lt;code&gt;proxy&lt;/code&gt;, &lt;code&gt;VPN&lt;/code&gt;, dial-up pool, cloud phone, or hijacked shared proxy) and was &lt;strong&gt;not&lt;/strong&gt; in the live 7-day risk mirror. Addresses that go quiet and return can be counted again. The page describes what &lt;strong&gt;86&lt;/strong&gt; upstream collection sources delivered — not the total number of proxies in the world. An address is counted once per window. No address lists are published.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why 2.4% still is not "mostly datacenter"
&lt;/h2&gt;

&lt;p&gt;If your defence is mainly an ASN or hosting blocklist, you are still looking past most of the problem. Residential broadband alone is &lt;strong&gt;78.7%&lt;/strong&gt; of this week's capture window. Mobile adds another &lt;strong&gt;13.4%&lt;/strong&gt;. Together that is &lt;strong&gt;92%&lt;/strong&gt; of newly seen proxy exits sitting in access networks where real customers also live.&lt;/p&gt;

&lt;p&gt;The datacenter row moved from &lt;strong&gt;2.2%&lt;/strong&gt; in the last two posted recaps (5–11 Nov and 9–15 Nov) to &lt;strong&gt;2.4%&lt;/strong&gt; here, still close to late October's &lt;strong&gt;2.0–2.2%&lt;/strong&gt; band after the brief rise to &lt;strong&gt;3.6%&lt;/strong&gt; (29 Oct–4 Nov) and &lt;strong&gt;3.2%&lt;/strong&gt; (4–10 Nov). A hosting-only blocklist still misses most of the supply-side traffic that shows up as residential or mobile exits.&lt;/p&gt;

&lt;p&gt;Usage type answers "what kind of network is this," not "is this address currently sold as an exit." Rotating residential proxies and dial-up pools are designed to sit in DYN / MOB space. Detecting them needs supply-side observation with an age attached, not a lasting network-type label. See &lt;a href="https://ip99.com/en/rotating-residential-proxies?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;rotating residential proxies&lt;/a&gt; and &lt;a href="https://ip99.com/en/detect-proxy-vpn?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;detect proxy / VPN&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  Which networks show up most
&lt;/h2&gt;

&lt;p&gt;Top ASNs in this week's capture window (by distinct address count):&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;ASN&lt;/th&gt;
&lt;th&gt;Addresses&lt;/th&gt;
&lt;th&gt;Share&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;AS4134&lt;/td&gt;
&lt;td&gt;947,700&lt;/td&gt;
&lt;td&gt;8.9%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;AS4837&lt;/td&gt;
&lt;td&gt;707,646&lt;/td&gt;
&lt;td&gt;6.6%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;AS9808&lt;/td&gt;
&lt;td&gt;358,103&lt;/td&gt;
&lt;td&gt;3.3%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;AS6057&lt;/td&gt;
&lt;td&gt;276,201&lt;/td&gt;
&lt;td&gt;2.6%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;AS55836&lt;/td&gt;
&lt;td&gt;259,403&lt;/td&gt;
&lt;td&gt;2.4%&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;AS4134, AS4837, and AS9808 remain the three largest contributors. The top-five change is fifth place: AS55836, eighth in the last two posted recaps (&lt;strong&gt;1.6%&lt;/strong&gt; on 5–11 Nov, &lt;strong&gt;1.8%&lt;/strong&gt; on 9–15 Nov) and fifth in the 29 Oct–4 Nov recap (2.2%), is fifth again at &lt;strong&gt;2.4%&lt;/strong&gt;. AS36903, fifth last week (&lt;strong&gt;2.2%&lt;/strong&gt;), sits seventh (&lt;strong&gt;2.0%&lt;/strong&gt;). Large access networks appearing here are a clue about where proxy supply is being sold from — not a claim that every customer on those networks is a proxy. Ban a whole ASN and you take out legitimate home and mobile users. Prefer per-address evidence. ASN pages: &lt;a href="https://ip99.com/en/asn/?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;ip99.com/en/asn/&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  Day-to-day shape
&lt;/h2&gt;

&lt;p&gt;New-to-mirror captures that day, and how many were high-risk at capture:&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Day (UTC)&lt;/th&gt;
&lt;th&gt;Captured&lt;/th&gt;
&lt;th&gt;High-risk at capture&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-12&lt;/td&gt;
&lt;td&gt;1,576,485&lt;/td&gt;
&lt;td&gt;1,000,400&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-13&lt;/td&gt;
&lt;td&gt;1,606,501&lt;/td&gt;
&lt;td&gt;994,471&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-14&lt;/td&gt;
&lt;td&gt;1,422,246&lt;/td&gt;
&lt;td&gt;896,773&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-15&lt;/td&gt;
&lt;td&gt;1,253,419&lt;/td&gt;
&lt;td&gt;804,951&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-16&lt;/td&gt;
&lt;td&gt;1,691,075&lt;/td&gt;
&lt;td&gt;1,032,943&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-17&lt;/td&gt;
&lt;td&gt;1,596,167&lt;/td&gt;
&lt;td&gt;983,685&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-18&lt;/td&gt;
&lt;td&gt;1,560,001&lt;/td&gt;
&lt;td&gt;1,003,907&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;Volume was up on the official comparison (&lt;strong&gt;+7%&lt;/strong&gt; vs 5–11 November). Upstream sources sat at &lt;strong&gt;86&lt;/strong&gt;, up from &lt;strong&gt;60&lt;/strong&gt; in the last posted recap — a large jump the page reports, not a cause this article will invent. Daily capture stayed in a &lt;strong&gt;1.25–1.69M&lt;/strong&gt; band, lightest on &lt;strong&gt;15 Nov&lt;/strong&gt; (&lt;strong&gt;1,253,419&lt;/strong&gt;) and heaviest on &lt;strong&gt;16 Nov&lt;/strong&gt; (&lt;strong&gt;1,691,075&lt;/strong&gt;). That is still not a shape that makes a static list look "caught up."&lt;/p&gt;

&lt;h2&gt;
  
  
  How to use this without over-blocking
&lt;/h2&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Branch on evidence, not network type.&lt;/strong&gt; Residential is where your customers live; do not hard-block on DYN / MOB alone.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Step up, don't stop.&lt;/strong&gt; A fresh proxy-class signal on signup, a coupon claim, or a login from a new device justifies a second factor or a delay — not an immediate ban of a whole building's shared egress.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Read the age.&lt;/strong&gt; &lt;code&gt;evidence_state: "stale"&lt;/code&gt; means the address was in supply before and is not now; treat it as a weak prior. Re-query at decision time rather than caching a verdict for days. Score &lt;code&gt;0&lt;/code&gt; with &lt;code&gt;evidence_state: "none"&lt;/code&gt; is not a clean bill of health — it means no current verifiable evidence is on file.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Keep the timestamp.&lt;/strong&gt; Store &lt;code&gt;computed_at&lt;/code&gt; with the decision so a disputed block can be traced back to when the observation was made.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;IP99 by ThreatHunter publishes the free weekly aggregate so you can see the mix without buying a feed. The score on a single lookup is the freshness and strength of verifiable risk evidence — not a probability that the address "is an attacker." Do not treat IP99 scores as identical to every other ThreatHunter dataset you may have seen elsewhere; this article only claims what the public proxy-exits page and free &lt;code&gt;/v1/ip&lt;/code&gt; document.&lt;/p&gt;

&lt;h2&gt;
  
  
  Check one yourself
&lt;/h2&gt;

&lt;p&gt;Free, no signup:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/8.8.8.8
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Look at &lt;code&gt;evidence_state&lt;/code&gt; (&lt;code&gt;active&lt;/code&gt; / &lt;code&gt;stale&lt;/code&gt; / &lt;code&gt;none&lt;/code&gt;), &lt;code&gt;risk.score&lt;/code&gt;, and the proxy-class signals. Verified 19 November 2026, 13:12 UTC: this public DNS address still shows a &lt;code&gt;hosting&lt;/code&gt; network attribute with score &lt;code&gt;0&lt;/code&gt;, &lt;code&gt;evidence_state: "none"&lt;/code&gt;, and &lt;code&gt;usage_type&lt;/code&gt; &lt;code&gt;IDC&lt;/code&gt; — a network fact, not a proxy verdict. The same lookup is available to agents as an MCP tool at &lt;code&gt;https://ip99.com/mcp&lt;/code&gt; (&lt;code&gt;ip_risk_lookup&lt;/code&gt;). Field notes live on the &lt;a href="https://ip99.com/api?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;API docs&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;Try it on &lt;a href="https://ip99.com?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;ip99.com&lt;/a&gt;, keep an eye on &lt;a href="https://ip99.com/en/proxy-exits?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;this week's proxy exits&lt;/a&gt;, and wire the free OpenAPI at &lt;a href="https://ip99.com/openapi.json?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;ip99.com/openapi.json&lt;/a&gt; for Postman or codegen.&lt;/p&gt;

</description>
      <category>networking</category>
      <category>security</category>
      <category>api</category>
      <category>cybersecurity</category>
    </item>
    <item>
      <title>An IP Timezone Is Not a Traveler's Clock: Reading geo.tz on IP99</title>
      <dc:creator>IP99</dc:creator>
      <pubDate>Wed, 18 Nov 2026 13:16:21 +0000</pubDate>
      <link>https://dev.to/ip99/an-ip-timezone-is-not-a-travelers-clock-reading-geotz-on-ip99-3jmh</link>
      <guid>https://dev.to/ip99/an-ip-timezone-is-not-a-travelers-clock-reading-geotz-on-ip99-3jmh</guid>
      <description>&lt;p&gt;An IP lookup that returns a time zone can feel like you have the user's clock. You do not. &lt;a href="https://ip99.com" rel="noopener noreferrer"&gt;IP99&lt;/a&gt; sets &lt;code&gt;geo.tz&lt;/code&gt; from the geolocation of the address block: an IANA time zone name the address is expected to be in, not a reading from the device, the browser, or NTP. A rule that compares a client's timezone to &lt;code&gt;geo.tz&lt;/code&gt; and fires on any mismatch is comparing a user setting to the documentation of a network. The same free, no-key call that returns risk also returns &lt;code&gt;geo&lt;/code&gt;. JSON samples were verified on 18 November 2026 UTC, using the free profile only. Notes: &lt;a href="https://ip99.com" rel="noopener noreferrer"&gt;IP99&lt;/a&gt;, &lt;a href="https://ip99.com/api" rel="noopener noreferrer"&gt;API&lt;/a&gt;, &lt;a href="https://ip99.com/en/ip-location" rel="noopener noreferrer"&gt;IP location&lt;/a&gt;, &lt;a href="https://ip99.com/en/free-ip-geolocation-risk-api" rel="noopener noreferrer"&gt;free geolocation and risk API&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  One curl, the tz field
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/8.8.8.8
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;No API key, no registration. Verified 18 November 2026 UTC (&lt;code&gt;computed_at&lt;/code&gt; &lt;code&gt;2026-11-18T13:11:20Z&lt;/code&gt;):&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="nl"&gt;"ip"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"8.8.8.8"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"computed_at"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"2026-11-18T13:11:20Z"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"evidence_state"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"geo"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
    &lt;span class="nl"&gt;"continent"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"NA"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"country"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"US"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"region"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"California"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"city"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Mountain View"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"district"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;""&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"lat"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mf"&gt;37.38605&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"lon"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mf"&gt;-122.08385&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"tz"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"America/Los_Angeles"&lt;/span&gt;
  &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"network"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"asn"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;15169&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"usage_type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"IDC"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"risk"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"score"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;0&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"level"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"signals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"hosting"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;&lt;code&gt;geo.tz&lt;/code&gt; is an IANA name, here &lt;code&gt;America/Los_Angeles&lt;/code&gt;. The six string keys — &lt;code&gt;continent&lt;/code&gt;, &lt;code&gt;country&lt;/code&gt;, &lt;code&gt;region&lt;/code&gt;, &lt;code&gt;city&lt;/code&gt;, &lt;code&gt;district&lt;/code&gt;, &lt;code&gt;tz&lt;/code&gt; — are always present on a successful lookup. A name the database does not have is an empty string, never a missing key and never a &lt;code&gt;0&lt;/code&gt;. Official field docs: &lt;code&gt;tz&lt;/code&gt; is empty only for a rare block with no timezone in the database.&lt;/p&gt;

&lt;p&gt;&lt;code&gt;hosting&lt;/code&gt; and &lt;code&gt;usage_type: "IDC"&lt;/code&gt; describe the network scene, not a fraud verdict — official guidance is not to block on &lt;code&gt;hosting&lt;/code&gt; or &lt;code&gt;mobile&lt;/code&gt; alone. Score 0 with &lt;code&gt;evidence_state: "none"&lt;/code&gt; means IP99 holds no current risk evidence, not that the address is safe. &lt;code&gt;risk.score&lt;/code&gt; is the freshness and strength of verifiable evidence, not attacker probability. Google Public DNS is anycast; Los Angeles time is a reasonable zone for the registered block, not the clock of the person who just logged in.&lt;/p&gt;

&lt;h2&gt;
  
  
  Compare the zone, not only the city
&lt;/h2&gt;

&lt;p&gt;Official guidance on &lt;a href="https://ip99.com/en/ip-location" rel="noopener noreferrer"&gt;how IP location works&lt;/a&gt; already treats the timezone as a check, not as a second pin on the map:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;Check for a VPN or proxy first. It is the most common reason a location "looks wrong," and the fix is not in the geolocation database — it is in &lt;code&gt;risk.signals&lt;/code&gt;.&lt;/li&gt;
&lt;li&gt;Compare &lt;code&gt;geo.tz&lt;/code&gt;, not just the city. A wrong city with a correct time zone usually means "registered at the regional equipment, used elsewhere." A timezone that is wrong as well is a stronger hint of a genuinely stale record.&lt;/li&gt;
&lt;li&gt;Then check &lt;code&gt;network.usage_type&lt;/code&gt;. &lt;code&gt;MOB&lt;/code&gt; or &lt;code&gt;GTW&lt;/code&gt; means the address is shared, and its city is approximate by construction.&lt;/li&gt;
&lt;li&gt;Then treat the result as an estimate. Use city-level location to pick a language, a currency, or a legal region — not to decide that a specific person is in a specific place.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;That is the point of &lt;code&gt;geo.tz&lt;/code&gt;: ask whether the block is still in the right zone, not whether a laptop set to UTC, a traveler who left the timezone unchanged, or a VPN exit should be rejected.&lt;/p&gt;

&lt;p&gt;The geolocation database is IP99's own. As of data version &lt;code&gt;20261118&lt;/code&gt; it covers &lt;strong&gt;244&lt;/strong&gt; countries and regions for IPv4 block attribution, operator, ASN, and usage type, and it is rebuilt daily (geo &lt;code&gt;20261118&lt;/code&gt;; risk &lt;code&gt;20261118211156&lt;/code&gt; the same day). The risk database is separate and updates every 10 seconds.&lt;/p&gt;

&lt;h2&gt;
  
  
  Two addresses, one ASN, two IANA zones
&lt;/h2&gt;

&lt;p&gt;Anycast and sibling addresses in the same network do not share a single clock. Verified the same day:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/1.1.1.1
curl https://ip99.com/v1/ip/1.0.0.1
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;&lt;code&gt;1.1.1.1&lt;/code&gt; (&lt;code&gt;computed_at&lt;/code&gt; &lt;code&gt;2026-11-18T13:11:24Z&lt;/code&gt;):&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="nl"&gt;"ip"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"1.1.1.1"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"computed_at"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"2026-11-18T13:11:24Z"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"evidence_state"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"geo"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"continent"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"OC"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"country"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"AU"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"region"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Queensland"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"city"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Brisbane"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"district"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;""&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"lat"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mf"&gt;-27.4679&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"lon"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mf"&gt;153.0281&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"tz"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Australia/Brisbane"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"network"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"asn"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;13335&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"usage_type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"IDC"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"risk"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"score"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;0&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"level"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"signals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"hosting"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;&lt;code&gt;1.0.0.1&lt;/code&gt; (&lt;code&gt;computed_at&lt;/code&gt; &lt;code&gt;2026-11-18T13:11:32Z&lt;/code&gt;):&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="nl"&gt;"ip"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"1.0.0.1"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"computed_at"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"2026-11-18T13:11:32Z"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"evidence_state"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"geo"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"continent"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"OC"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"country"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"AU"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"region"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"New South Wales"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"city"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Sydney"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"district"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;""&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"lat"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mf"&gt;-33.86785&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"lon"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mf"&gt;151.20732&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"tz"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Australia/Sydney"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"network"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"asn"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;13335&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"usage_type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"IDC"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"risk"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"score"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;0&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"level"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"signals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"hosting"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Both are Cloudflare addresses in AS13335, both &lt;code&gt;IDC&lt;/code&gt; with &lt;code&gt;hosting&lt;/code&gt;, both in Australia — and they are not the same IANA zone. &lt;code&gt;Australia/Brisbane&lt;/code&gt; and &lt;code&gt;Australia/Sydney&lt;/code&gt; do not always share an offset: Queensland does not observe daylight saving; New South Wales does. Collapsing the country into one "AEST" zone would already be wrong; treating either city as "where the user is" would be worse. These are registration and routing facts about the blocks. The resolver is anycast; Brisbane or Sydney is not the node that answered you, and it is not a traveler's clock.&lt;/p&gt;

&lt;p&gt;IPv4 and IPv6 exits can disagree as well. A lookup of &lt;code&gt;2606:4700:4700::1111&lt;/code&gt; the same day (&lt;code&gt;computed_at&lt;/code&gt; &lt;code&gt;2026-11-18T13:11:45Z&lt;/code&gt;) returned San Francisco and &lt;code&gt;America/Los_Angeles&lt;/code&gt;, not &lt;code&gt;Australia/Brisbane&lt;/code&gt;. A dual-stack client may present a different egress family on different requests, so a "client timezone versus IP timezone" rule can fire on one family and stay quiet on the other, for the same person.&lt;/p&gt;

&lt;h2&gt;
  
  
  When a mismatch is the expected answer
&lt;/h2&gt;

&lt;p&gt;A consumer address still describes a block, not a person. Official docs use &lt;code&gt;27.22.66.200&lt;/code&gt; as the China example. Live on 18 November 2026 UTC:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/27.22.66.200
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;





&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="nl"&gt;"ip"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"27.22.66.200"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"computed_at"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"2026-11-18T13:11:27Z"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"evidence_state"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"active"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"geo"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"continent"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"AS"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"country"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"CN"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"region"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Hubei"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"city"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Xiangyang"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"district"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Fancheng"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"lat"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mf"&gt;32.04482&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"lon"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mf"&gt;112.13546&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"tz"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Asia/Shanghai"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"network"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"asn"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;4134&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"usage_type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"DYN"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"risk"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"score"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;10&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"level"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"low"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"signals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"dialup_pool"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"hijacked_proxy"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"proxy"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;&lt;code&gt;Asia/Shanghai&lt;/code&gt; matches the country and the region. That does not mean a specific person is in Fancheng, and a browser set to &lt;code&gt;Asia/Shanghai&lt;/code&gt; does not "confirm" the login. This address also carries current risk evidence (&lt;code&gt;dialup_pool&lt;/code&gt;, &lt;code&gt;hijacked_proxy&lt;/code&gt;, &lt;code&gt;proxy&lt;/code&gt;) at score 10 / &lt;code&gt;low&lt;/code&gt; — a separate reading from &lt;code&gt;geo&lt;/code&gt;. Official guidance: check those signals before you argue with the city or the zone, and do not hard-ban an account on &lt;code&gt;hijacked_proxy&lt;/code&gt; alone; the holder is often a victim.&lt;/p&gt;

&lt;p&gt;A timezone mismatch is often expected. A VPN or proxy exit is reported in the exit's zone — look at &lt;code&gt;risk.signals&lt;/code&gt; first. &lt;code&gt;MOB&lt;/code&gt; or a shared &lt;code&gt;GTW&lt;/code&gt; address serves a wide area. Anycast, hosting, and corporate headquarters egress describe the server or the registered block, rarely the user. And the client timezone is a setting: laptops stay on the home zone during travel, some systems are set to UTC, and a VM inherits the image default.&lt;/p&gt;

&lt;p&gt;None of that is a reason to skip the field. It is a reason not to treat disagreement as proof of account takeover. If the city looks off but &lt;code&gt;geo.tz&lt;/code&gt; is consistent with the region, prefer "registered elsewhere, used in-zone" over "stale data." If the zone is also implausible, discount the record and add a check — a step-up challenge, not an automatic ban. Combine it with evidence state and signals.&lt;/p&gt;

&lt;h2&gt;
  
  
  Try it
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/8.8.8.8
curl https://ip99.com/v1/ip/1.1.1.1
curl https://ip99.com/v1/ip/1.0.0.1
curl https://ip99.com/v1/ip/2606:4700:4700::1111
curl https://ip99.com/v1/ip/27.22.66.200
curl https://ip99.com/v1/me
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Anonymous quota is &lt;strong&gt;500 lookups/day&lt;/strong&gt; per client IP (60/min); a free logged-in key raises it to &lt;strong&gt;5,000/day&lt;/strong&gt;. Check remaining with &lt;code&gt;GET https://ip99.com/v1/me&lt;/code&gt;. OpenAPI: &lt;a href="https://ip99.com/openapi.json" rel="noopener noreferrer"&gt;https://ip99.com/openapi.json&lt;/a&gt;. Pricing: &lt;a href="https://ip99.com/en/pricing" rel="noopener noreferrer"&gt;https://ip99.com/en/pricing&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;Do &lt;strong&gt;not&lt;/strong&gt; treat IP99 scores or geo as identical to other ThreatHunter datasets. This article only claims what free &lt;code&gt;/v1/ip&lt;/code&gt; and the cited product pages document.&lt;/p&gt;

&lt;p&gt;IP99 is a ThreatHunter product. An IP timezone is the IANA zone of a network block, not a traveler's clock and not a person. Start at &lt;a href="https://ip99.com" rel="noopener noreferrer"&gt;https://ip99.com&lt;/a&gt;; the field contract is on &lt;a href="https://ip99.com/en/ip-location" rel="noopener noreferrer"&gt;IP location&lt;/a&gt;, the &lt;a href="https://ip99.com/en/free-ip-geolocation-risk-api" rel="noopener noreferrer"&gt;free geolocation and risk API&lt;/a&gt;, and &lt;a href="https://ip99.com/api" rel="noopener noreferrer"&gt;https://ip99.com/api&lt;/a&gt;.&lt;/p&gt;

</description>
      <category>cybersecurity</category>
      <category>api</category>
      <category>geolocation</category>
      <category>networking</category>
    </item>
    <item>
      <title>IP Location Is Not a Street Address: Reading geo Fields on IP99</title>
      <dc:creator>IP99</dc:creator>
      <pubDate>Tue, 17 Nov 2026 13:19:45 +0000</pubDate>
      <link>https://dev.to/ip99/ip-location-is-not-a-street-address-reading-geo-fields-on-ip99-1acb</link>
      <guid>https://dev.to/ip99/ip-location-is-not-a-street-address-reading-geo-fields-on-ip99-1acb</guid>
      <description>&lt;p&gt;An IP lookup that returns a city can feel like a pin on a map. It is not. &lt;a href="https://ip99.com" rel="noopener noreferrer"&gt;IP99&lt;/a&gt; derives a location from registry and routing data for the address block, not from measuring a device. Country is usually reliable; region and city are usually right for consumer broadband; a street address is impossible. The same free, no-key call that returns risk also returns &lt;code&gt;geo&lt;/code&gt;. JSON samples were verified on 17 November 2026 UTC, using the free profile only. Notes: &lt;a href="https://ip99.com" rel="noopener noreferrer"&gt;IP99&lt;/a&gt;, &lt;a href="https://ip99.com/api" rel="noopener noreferrer"&gt;API&lt;/a&gt;, &lt;a href="https://ip99.com/en/ip-location" rel="noopener noreferrer"&gt;IP location&lt;/a&gt;, &lt;a href="https://ip99.com/en/free-ip-geolocation-risk-api" rel="noopener noreferrer"&gt;free geolocation and risk API&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  One curl, the whole geo object
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/8.8.8.8
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;No API key, no registration. Verified 17 November 2026 UTC (&lt;code&gt;computed_at&lt;/code&gt; &lt;code&gt;2026-11-17T13:11:28Z&lt;/code&gt;):&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="nl"&gt;"ip"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"8.8.8.8"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"computed_at"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"2026-11-17T13:11:28Z"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"evidence_state"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"geo"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
    &lt;span class="nl"&gt;"continent"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"NA"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"country"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"US"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"region"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"California"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"city"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Mountain View"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"district"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;""&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"lat"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mf"&gt;37.38605&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"lon"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mf"&gt;-122.08385&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"tz"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"America/Los_Angeles"&lt;/span&gt;
  &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"network"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"asn"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;15169&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"usage_type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"IDC"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"risk"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"score"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;0&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"level"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"signals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"hosting"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Read the fields as a contract, not as a street. &lt;code&gt;geo.country&lt;/code&gt; is an ISO 3166-1 alpha-2 code; &lt;code&gt;geo.region&lt;/code&gt; and &lt;code&gt;geo.city&lt;/code&gt; are Latin place names; &lt;code&gt;geo.tz&lt;/code&gt; is an IANA zone; &lt;code&gt;geo.continent&lt;/code&gt; is one of &lt;code&gt;AS&lt;/code&gt;, &lt;code&gt;EU&lt;/code&gt;, &lt;code&gt;NA&lt;/code&gt;, &lt;code&gt;SA&lt;/code&gt;, &lt;code&gt;AF&lt;/code&gt;, &lt;code&gt;OC&lt;/code&gt;, &lt;code&gt;AN&lt;/code&gt;. &lt;code&gt;geo.lat&lt;/code&gt; / &lt;code&gt;geo.lon&lt;/code&gt; are coordinates of the block, rounded to at most five decimal places, sent as a pair and omitted together when unknown — because &lt;code&gt;0,0&lt;/code&gt; is a real point in the Gulf of Guinea.&lt;/p&gt;

&lt;p&gt;The six string keys — &lt;code&gt;continent&lt;/code&gt;, &lt;code&gt;country&lt;/code&gt;, &lt;code&gt;region&lt;/code&gt;, &lt;code&gt;city&lt;/code&gt;, &lt;code&gt;district&lt;/code&gt;, &lt;code&gt;tz&lt;/code&gt; — are always present. A name the database does not have is an empty string, never a missing key and never a &lt;code&gt;0&lt;/code&gt;. That is why &lt;code&gt;district&lt;/code&gt; is &lt;code&gt;""&lt;/code&gt; here: the English-language place-name database covers countries other than China down to the city. One parser can handle every row.&lt;/p&gt;

&lt;p&gt;&lt;code&gt;hosting&lt;/code&gt; and &lt;code&gt;usage_type: "IDC"&lt;/code&gt; describe the network scene, not a fraud verdict — official guidance is not to block on &lt;code&gt;hosting&lt;/code&gt; or &lt;code&gt;mobile&lt;/code&gt; alone. Score 0 with &lt;code&gt;evidence_state: "none"&lt;/code&gt; means IP99 holds no current risk evidence, not that the address is safe. &lt;code&gt;risk.score&lt;/code&gt; is the freshness and strength of verifiable evidence, not attacker probability. Google Public DNS is anycast; Mountain View is a reasonable registration point for the block, not "the building this packet left."&lt;/p&gt;

&lt;h2&gt;
  
  
  Accuracy falls off as you zoom in
&lt;/h2&gt;

&lt;p&gt;Official guidance on &lt;a href="https://ip99.com/en/ip-location" rel="noopener noreferrer"&gt;how IP location works&lt;/a&gt; is a gradient, not a single accuracy number:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Country&lt;/strong&gt; — high confidence. Errors are rare and usually mean a whole block was moved, not that one address was misread.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Region&lt;/strong&gt; — good. Large ISPs often register every address to the region that holds their equipment.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;City&lt;/strong&gt; — statistical, not a fix. Consumer broadband usually lands in the right metropolitan area; mobile, corporate, and VPN ranges regularly show the wrong city or the carrier's headquarters.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;District&lt;/strong&gt; — only where IP99 has it. The value is non-empty for addresses in China; elsewhere the key is present as an empty string. Finer than city, coarser than a street.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;The geolocation database is IP99's own. As of data version &lt;code&gt;20261117&lt;/code&gt; it covers &lt;strong&gt;244&lt;/strong&gt; countries and regions for IPv4 block attribution, operator, ASN, and usage type, and it is rebuilt daily. The risk database is separate and updates every 10 seconds.&lt;/p&gt;

&lt;h2&gt;
  
  
  Anycast is not "where the user is"
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/1.1.1.1
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Verified the same day (&lt;code&gt;computed_at&lt;/code&gt; &lt;code&gt;2026-11-17T13:11:32Z&lt;/code&gt;):&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="nl"&gt;"ip"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"1.1.1.1"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"computed_at"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"2026-11-17T13:11:32Z"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"evidence_state"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"geo"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"continent"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"OC"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"country"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"AU"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"region"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Queensland"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"city"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Brisbane"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"district"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;""&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"lat"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mf"&gt;-27.4679&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"lon"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mf"&gt;153.0281&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"tz"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Australia/Brisbane"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"network"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"asn"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;13335&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"usage_type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"IDC"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"risk"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"score"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;0&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"level"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"signals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"hosting"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Cloudflare's public resolver is also anycast. The lookup reports Brisbane, Australia — a registration and routing fact about the block, not the city of the node that answered you, and not the city of a person using 1.1.1.1 as DNS. Using that city to pick a legal region, or to fire a "login from a new country" rule, would be reacting to the exit's documentation, not to a traveller.&lt;/p&gt;

&lt;p&gt;IPv6 uses the same endpoint and the same &lt;code&gt;geo&lt;/code&gt; fields. A lookup of &lt;code&gt;2001:4860:4860::8888&lt;/code&gt; the same day returned the same Mountain View record as &lt;code&gt;8.8.8.8&lt;/code&gt;. Coverage still depends on the registrations for that prefix; a newly allocated range may resolve only to a country.&lt;/p&gt;

&lt;h2&gt;
  
  
  District, when the database has it
&lt;/h2&gt;

&lt;p&gt;For an address in China the same three Latin name fields are filled, and &lt;code&gt;district&lt;/code&gt; is no longer empty. Official docs use &lt;code&gt;27.22.66.200&lt;/code&gt; as the example. Live on 17 November 2026 UTC (&lt;code&gt;computed_at&lt;/code&gt; &lt;code&gt;2026-11-17T13:11:34Z&lt;/code&gt;):&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/27.22.66.200
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;





&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="nl"&gt;"ip"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"27.22.66.200"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"computed_at"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"2026-11-17T13:11:34Z"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"evidence_state"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"active"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"geo"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"continent"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"AS"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"country"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"CN"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"region"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Hubei"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"city"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Xiangyang"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"district"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Fancheng"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"lat"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mf"&gt;32.04482&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"lon"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mf"&gt;112.13546&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"tz"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Asia/Shanghai"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"network"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"asn"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;4134&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"usage_type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"DYN"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"risk"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"score"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;15&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"level"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"low"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"signals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"dialup_pool"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"hijacked_proxy"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"proxy"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;&lt;code&gt;region&lt;/code&gt; / &lt;code&gt;city&lt;/code&gt; / &lt;code&gt;district&lt;/code&gt; here are Hubei / Xiangyang / Fancheng. Because the geolocation database is rebuilt daily, a published example can move by a district as the mapping is updated — treat the live response as the source, not a screenshot from last month. This address also carries current risk evidence (&lt;code&gt;dialup_pool&lt;/code&gt;, &lt;code&gt;hijacked_proxy&lt;/code&gt;, &lt;code&gt;proxy&lt;/code&gt;) at score 15 / &lt;code&gt;low&lt;/code&gt;, a separate reading from &lt;code&gt;geo&lt;/code&gt;. Official guidance: &lt;code&gt;hijacked_proxy&lt;/code&gt; often means the holder is a victim, so do not hard-ban the account on that signal alone. See &lt;a href="https://ip99.com/en/geo/cn" rel="noopener noreferrer"&gt;https://ip99.com/en/geo/cn&lt;/a&gt; for a country-level range view.&lt;/p&gt;

&lt;h2&gt;
  
  
  When the location looks wrong
&lt;/h2&gt;

&lt;p&gt;Nothing in the API is "broken" in the usual cases people report. A block can be registered in one city and used in another; the lookup reports the registration. A VPN or proxy replaces the egress, so the lookup reports where the exit is. Mobile and carrier-grade NAT share one address across a region (&lt;code&gt;MOB&lt;/code&gt; or &lt;code&gt;GTW&lt;/code&gt; means the city is approximate by construction). Corporate and satellite links can egress through a headquarters or a ground station. A datacenter address describes the server or anycast announcement, rarely the users.&lt;/p&gt;

&lt;p&gt;A practical order, from the official guide: check for a VPN or proxy first — it is the most common cause, and the fix is not in the geolocation database. Compare &lt;code&gt;geo.tz&lt;/code&gt;, not just the city: a wrong city with a correct time zone usually means "registered at the regional equipment, used elsewhere." Then look at &lt;code&gt;usage_type&lt;/code&gt;. Then treat the result as an estimate. Use city-level location to pick a language, a currency, or a legal region — not to decide that a specific person is in a specific place.&lt;/p&gt;

&lt;h2&gt;
  
  
  Try it
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/8.8.8.8
curl https://ip99.com/v1/ip/1.1.1.1
curl https://ip99.com/v1/ip/27.22.66.200
curl https://ip99.com/v1/me
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Anonymous quota is &lt;strong&gt;500 lookups/day&lt;/strong&gt; per client IP (60/min); a free logged-in key raises it to &lt;strong&gt;5,000/day&lt;/strong&gt;. Check remaining with &lt;code&gt;GET https://ip99.com/v1/me&lt;/code&gt; or the &lt;code&gt;X-Daily-*&lt;/code&gt; headers. OpenAPI: &lt;a href="https://ip99.com/openapi.json" rel="noopener noreferrer"&gt;https://ip99.com/openapi.json&lt;/a&gt;. Pricing: &lt;a href="https://ip99.com/en/pricing" rel="noopener noreferrer"&gt;https://ip99.com/en/pricing&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;Do &lt;strong&gt;not&lt;/strong&gt; treat IP99 scores or geo as identical to other ThreatHunter datasets — this article only claims what free &lt;code&gt;/v1/ip&lt;/code&gt; and the cited product pages document.&lt;/p&gt;

&lt;p&gt;IP99 is a ThreatHunter product. An IP location is an estimate for a network block, not a street address and not a person. Start at &lt;a href="https://ip99.com" rel="noopener noreferrer"&gt;https://ip99.com&lt;/a&gt;; the field contract is on &lt;a href="https://ip99.com/en/ip-location" rel="noopener noreferrer"&gt;IP location&lt;/a&gt;, the &lt;a href="https://ip99.com/en/free-ip-geolocation-risk-api" rel="noopener noreferrer"&gt;free geolocation and risk API&lt;/a&gt;, and &lt;a href="https://ip99.com/api" rel="noopener noreferrer"&gt;https://ip99.com/api&lt;/a&gt;.&lt;/p&gt;

</description>
      <category>networking</category>
      <category>cybersecurity</category>
      <category>api</category>
      <category>geolocation</category>
    </item>
    <item>
      <title>This Week's Proxy Exits: 9.9M New Addresses, Still Almost All Off Datacenter Space</title>
      <dc:creator>IP99</dc:creator>
      <pubDate>Mon, 16 Nov 2026 13:11:30 +0000</pubDate>
      <link>https://dev.to/ip99/this-weeks-proxy-exits-99m-new-addresses-still-almost-all-off-datacenter-space-4005</link>
      <guid>https://dev.to/ip99/this-weeks-proxy-exits-99m-new-addresses-still-almost-all-off-datacenter-space-4005</guid>
      <description>&lt;p&gt;Datacenter blocklists still look like the default proxy defence in a lot of fraud stacks. This week's capture window was softer on volume, but the mix did not flip: newly seen proxy exits were still almost all off datacenter space. The datacenter slice held at &lt;strong&gt;2.2%&lt;/strong&gt;, the same share as the last posted recap, still in the flat ~2% band of late October.&lt;/p&gt;

&lt;p&gt;Between &lt;strong&gt;9–15 November 2026 (UTC)&lt;/strong&gt;, the &lt;a href="https://ip99.com?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP99&lt;/a&gt; risk feed captured &lt;strong&gt;9,914,335&lt;/strong&gt; proxy exit addresses it had not seen in the previous 7 days (&lt;strong&gt;−8%&lt;/strong&gt; vs the previous 7 days). Source page: &lt;a href="https://ip99.com/en/proxy-exits?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;ip99.com/en/proxy-exits&lt;/a&gt; (generated 16 November 2026 UTC). The public page rolls daily over the last 7 complete UTC days; this is the window it reports today, continuing the series after the 5–11 November recap (the two windows overlap on 9–11 November).&lt;/p&gt;

&lt;p&gt;Against that last posted week (&lt;strong&gt;10,026,346&lt;/strong&gt; addresses), volume is only down about &lt;strong&gt;−1.1%&lt;/strong&gt;. The page's official &lt;strong&gt;−8%&lt;/strong&gt; compares 9–15 November with 2–8 November, not with the overlapping posted window. A large part of that official step-down is compositional: the comparison window still includes the heavier days of &lt;strong&gt;3 November&lt;/strong&gt; (&lt;strong&gt;1,642,470&lt;/strong&gt;) and &lt;strong&gt;4 November&lt;/strong&gt; (&lt;strong&gt;1,752,678&lt;/strong&gt;), which had already rolled out of the last posted recap. These figures describe what entered the 7-day mirror, not worldwide proxy supply.&lt;/p&gt;

&lt;h2&gt;
  
  
  What those addresses sit on
&lt;/h2&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Usage type&lt;/th&gt;
&lt;th&gt;Addresses&lt;/th&gt;
&lt;th&gt;Share&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;Residential broadband&lt;/td&gt;
&lt;td&gt;7,951,392&lt;/td&gt;
&lt;td&gt;80.2%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Mobile network&lt;/td&gt;
&lt;td&gt;1,245,122&lt;/td&gt;
&lt;td&gt;12.6%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Gateway / NAT egress&lt;/td&gt;
&lt;td&gt;425,455&lt;/td&gt;
&lt;td&gt;4.3%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Datacenter&lt;/td&gt;
&lt;td&gt;214,585&lt;/td&gt;
&lt;td&gt;2.2%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Education network&lt;/td&gt;
&lt;td&gt;49,101&lt;/td&gt;
&lt;td&gt;0.5%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Unclassified&lt;/td&gt;
&lt;td&gt;28,680&lt;/td&gt;
&lt;td&gt;0.3%&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;About &lt;strong&gt;63%&lt;/strong&gt; were rated high-risk at capture; the rest still carried a proxy-class signal at a lower risk level.&lt;/p&gt;

&lt;p&gt;IP99 counts an address as "captured" when it enters the feed with a proxy-class signal (&lt;code&gt;proxy&lt;/code&gt;, &lt;code&gt;VPN&lt;/code&gt;, dial-up pool, cloud phone, or hijacked shared proxy) and was &lt;strong&gt;not&lt;/strong&gt; in the live 7-day risk mirror. Addresses that go quiet and return can be counted again. The page describes what &lt;strong&gt;60&lt;/strong&gt; upstream collection sources delivered — not the total number of proxies in the world. An address is counted once per window. No address lists are published.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why 2.2% still is not "mostly datacenter"
&lt;/h2&gt;

&lt;p&gt;If your defence is mainly an ASN or hosting blocklist, you are still mostly looking past the problem. Residential broadband alone is &lt;strong&gt;80.2%&lt;/strong&gt; of this week's capture window. Mobile adds another &lt;strong&gt;12.6%&lt;/strong&gt;. Together that is &lt;strong&gt;93%&lt;/strong&gt; of newly seen proxy exits sitting in access networks where real customers also live.&lt;/p&gt;

&lt;p&gt;The datacenter row is unchanged at &lt;strong&gt;2.2%&lt;/strong&gt; versus the last posted recap (5–11 Nov), after easing from &lt;strong&gt;3.2%&lt;/strong&gt; (4–10 Nov) and &lt;strong&gt;3.6%&lt;/strong&gt; (29 Oct–4 Nov) toward the &lt;strong&gt;2.0%&lt;/strong&gt; of late October. A hosting-only blocklist still misses most of the supply-side traffic that shows up as residential or mobile exits.&lt;/p&gt;

&lt;p&gt;Residential share eased slightly from &lt;strong&gt;80.7%&lt;/strong&gt; to &lt;strong&gt;80.2%&lt;/strong&gt;, with mobile (&lt;strong&gt;12.2%&lt;/strong&gt; → &lt;strong&gt;12.6%&lt;/strong&gt;) and gateway / NAT egress (&lt;strong&gt;4.0%&lt;/strong&gt; → &lt;strong&gt;4.3%&lt;/strong&gt;) taking the difference — small mix moves inside the same access-network story.&lt;/p&gt;

&lt;p&gt;Usage type answers "what kind of network is this," not "is this address currently sold as an exit." Rotating residential proxies and dial-up pools are designed to sit in DYN / MOB space. Detecting them needs supply-side observation with an age attached, not a permanent network-type label. Product notes: &lt;a href="https://ip99.com/en/rotating-residential-proxies?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;rotating residential proxies&lt;/a&gt;, &lt;a href="https://ip99.com/en/detect-proxy-vpn?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;detect proxy / VPN&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  Which networks show up most
&lt;/h2&gt;

&lt;p&gt;Top ASNs in this week's capture window (by distinct address count):&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;ASN&lt;/th&gt;
&lt;th&gt;Addresses&lt;/th&gt;
&lt;th&gt;Share&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;AS4134&lt;/td&gt;
&lt;td&gt;914,073&lt;/td&gt;
&lt;td&gt;9.2%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;AS4837&lt;/td&gt;
&lt;td&gt;691,730&lt;/td&gt;
&lt;td&gt;7.0%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;AS9808&lt;/td&gt;
&lt;td&gt;340,319&lt;/td&gt;
&lt;td&gt;3.4%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;AS6057&lt;/td&gt;
&lt;td&gt;255,175&lt;/td&gt;
&lt;td&gt;2.6%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;AS36903&lt;/td&gt;
&lt;td&gt;213,898&lt;/td&gt;
&lt;td&gt;2.2%&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;AS4134, AS4837, and AS9808 remain the three largest contributors. The top five is unchanged from the 5–11 November recap: AS6057 still fourth, AS36903 still fifth (share &lt;strong&gt;2.0%&lt;/strong&gt; → &lt;strong&gt;2.2%&lt;/strong&gt;). AS55836, fifth in the 29 Oct–4 Nov recap (2.2%), remains eighth this week (&lt;strong&gt;1.8%&lt;/strong&gt;, up from &lt;strong&gt;1.6%&lt;/strong&gt; last time). Large access networks appearing here are a clue about where proxy supply is being sold from — not a claim that every customer on those networks is a proxy. Ban a whole ASN and you take out legitimate home and mobile users. Prefer per-address evidence. ASN pages: &lt;a href="https://ip99.com/en/asn/?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;ip99.com/en/asn/&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  Day-to-day shape
&lt;/h2&gt;

&lt;p&gt;New-to-mirror captures that day, and how many were high-risk at capture:&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Day (UTC)&lt;/th&gt;
&lt;th&gt;Captured&lt;/th&gt;
&lt;th&gt;High-risk at capture&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-09&lt;/td&gt;
&lt;td&gt;1,371,503&lt;/td&gt;
&lt;td&gt;883,005&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-10&lt;/td&gt;
&lt;td&gt;1,467,002&lt;/td&gt;
&lt;td&gt;940,244&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-11&lt;/td&gt;
&lt;td&gt;1,217,179&lt;/td&gt;
&lt;td&gt;695,544&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-12&lt;/td&gt;
&lt;td&gt;1,576,485&lt;/td&gt;
&lt;td&gt;1,000,400&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-13&lt;/td&gt;
&lt;td&gt;1,606,501&lt;/td&gt;
&lt;td&gt;994,471&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-14&lt;/td&gt;
&lt;td&gt;1,422,246&lt;/td&gt;
&lt;td&gt;896,773&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-15&lt;/td&gt;
&lt;td&gt;1,253,419&lt;/td&gt;
&lt;td&gt;804,951&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;Volume was down on the official comparison (&lt;strong&gt;−8%&lt;/strong&gt; vs 2–8 November); upstream sources sat at &lt;strong&gt;60&lt;/strong&gt; (56 in the last posted recap). Daily capture stayed in a &lt;strong&gt;1.22–1.61M&lt;/strong&gt; band, lightest on &lt;strong&gt;11 Nov&lt;/strong&gt; (&lt;strong&gt;1,217,179&lt;/strong&gt;) and &lt;strong&gt;15 Nov&lt;/strong&gt; (&lt;strong&gt;1,253,419&lt;/strong&gt;), heaviest on &lt;strong&gt;13 Nov&lt;/strong&gt; (&lt;strong&gt;1,606,501&lt;/strong&gt;). That is still not a shape that makes a static list look "caught up."&lt;/p&gt;

&lt;h2&gt;
  
  
  How to use this without over-blocking
&lt;/h2&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Branch on evidence, not network type.&lt;/strong&gt; Residential is where your customers live; do not hard-block on DYN / MOB alone.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Step up, don't stop.&lt;/strong&gt; A fresh proxy-class signal on signup, a coupon claim, or a login from a new device justifies a second factor or a delay — not an immediate ban of a whole building's shared egress.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Read the age.&lt;/strong&gt; &lt;code&gt;evidence_state: "stale"&lt;/code&gt; means the address was in supply before and is not now; treat it as a weak prior. Re-query at decision time rather than caching a verdict for days. Score &lt;code&gt;0&lt;/code&gt; with &lt;code&gt;evidence_state: "none"&lt;/code&gt; is not a clean bill of health — it means no current verifiable evidence is on file.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Keep the timestamp.&lt;/strong&gt; Store &lt;code&gt;computed_at&lt;/code&gt; with the decision so a disputed block can be traced back to when the observation was made.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;IP99 by ThreatHunter publishes the free weekly aggregate so you can see the mix without buying a feed. The score on a single lookup is the freshness and strength of verifiable risk evidence — not a probability that the address "is an attacker." Do not treat IP99 scores as identical to every other ThreatHunter dataset you may have seen elsewhere; this article only claims what the public proxy-exits page and free &lt;code&gt;/v1/ip&lt;/code&gt; document.&lt;/p&gt;

&lt;h2&gt;
  
  
  Check one yourself
&lt;/h2&gt;

&lt;p&gt;Free, no signup:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/8.8.8.8
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Look at &lt;code&gt;evidence_state&lt;/code&gt; (&lt;code&gt;active&lt;/code&gt; / &lt;code&gt;stale&lt;/code&gt; / &lt;code&gt;none&lt;/code&gt;), &lt;code&gt;risk.score&lt;/code&gt;, and the proxy-class signals. Verified 16 November 2026, 13:05 UTC: public DNS infrastructure here still shows a &lt;code&gt;hosting&lt;/code&gt; network attribute with score &lt;code&gt;0&lt;/code&gt;, &lt;code&gt;evidence_state: "none"&lt;/code&gt;, and &lt;code&gt;usage_type&lt;/code&gt; &lt;code&gt;IDC&lt;/code&gt; — a network fact, not a proxy verdict. The same lookup is available to agents as an MCP tool at &lt;code&gt;https://ip99.com/mcp&lt;/code&gt; (&lt;code&gt;ip_risk_lookup&lt;/code&gt;). Field notes live on the &lt;a href="https://ip99.com/api?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;API docs&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;Try it on &lt;a href="https://ip99.com?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;ip99.com&lt;/a&gt;, keep an eye on &lt;a href="https://ip99.com/en/proxy-exits?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;this week's proxy exits&lt;/a&gt;, and wire the free OpenAPI at &lt;a href="https://ip99.com/openapi.json?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;ip99.com/openapi.json&lt;/a&gt; for Postman or codegen.&lt;/p&gt;

</description>
      <category>security</category>
      <category>api</category>
      <category>cybersecurity</category>
      <category>networking</category>
    </item>
    <item>
      <title>Check the Exit the Internet Sees: Looking Up Your Own IP on IP99</title>
      <dc:creator>IP99</dc:creator>
      <pubDate>Sun, 15 Nov 2026 13:19:46 +0000</pubDate>
      <link>https://dev.to/ip99/check-the-exit-the-internet-sees-looking-up-your-own-ip-on-ip99-1l7n</link>
      <guid>https://dev.to/ip99/check-the-exit-the-internet-sees-looking-up-your-own-ip-on-ip99-1l7n</guid>
      <description>&lt;p&gt;When a request is blocked, a model refuses a prompt, or a "login from a new country" email fires, the first question is usually "what address did they see?" The address printed on your laptop is often the wrong answer. &lt;code&gt;192.168.1.1&lt;/code&gt; and &lt;code&gt;10.0.0.1&lt;/code&gt; are private; they are reused on every LAN in the world and are never routed on the public internet. What a remote service sees is the public egress of your NAT, VPN, office gateway, or CI runner.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://ip99.com/?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP99&lt;/a&gt; by ThreatHunter has a no-key call for exactly that: &lt;code&gt;GET /v1/ip&lt;/code&gt; with no address in the path. It returns the caller's public IP plus the same risk, network, and geolocation profile you get when you look up someone else. The HTML version lives at &lt;a href="https://ip99.com/en/what-is-my-ip?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;What is my IP address?&lt;/a&gt;. This walkthrough uses the free profile only. JSON samples were verified on 15 November 2026 UTC. Notes: &lt;a href="https://ip99.com/?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP99&lt;/a&gt;, &lt;a href="https://ip99.com/api?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;API&lt;/a&gt;, &lt;a href="https://ip99.com/en/what-is-my-ip?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;What is my IP&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  The one-line lookup
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;No API key, no registration. The path is &lt;code&gt;/v1/ip&lt;/code&gt;, not &lt;code&gt;/v1/ip/{ip}&lt;/code&gt;. The response shape matches a normal lookup: &lt;code&gt;ip&lt;/code&gt;, &lt;code&gt;computed_at&lt;/code&gt;, &lt;code&gt;evidence_state&lt;/code&gt;, &lt;code&gt;risk&lt;/code&gt;, &lt;code&gt;network&lt;/code&gt;, &lt;code&gt;geo&lt;/code&gt;.&lt;/p&gt;

&lt;p&gt;A real call from this article's publishing environment on 15 November 2026 UTC (&lt;code&gt;computed_at&lt;/code&gt; &lt;code&gt;2026-11-15T13:14:21Z&lt;/code&gt;):&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="nl"&gt;"ip"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"104.28.245.205"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"computed_at"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"2026-11-15T13:14:21Z"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"evidence_state"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"geo"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
    &lt;span class="nl"&gt;"continent"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"AS"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"country"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"CN"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"region"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Henan"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"city"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Zhengzhou"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"district"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;""&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"lat"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mf"&gt;34.75798&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"lon"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mf"&gt;113.66541&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"tz"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Asia/Shanghai"&lt;/span&gt;
  &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"network"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"asn"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;13335&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"usage_type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"IDC"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"risk"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"score"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;0&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"level"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"signals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"hosting"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;The internet did not see a laptop's RFC1918 address. It saw a Cloudflare (AS13335) datacenter exit. &lt;code&gt;usage_type: "IDC"&lt;/code&gt; and &lt;code&gt;signals: ["hosting"]&lt;/code&gt; describe the network scene, not a fraud verdict — official guidance is not to block on &lt;code&gt;hosting&lt;/code&gt; or &lt;code&gt;mobile&lt;/code&gt; alone. Score 0 with &lt;code&gt;evidence_state: "none"&lt;/code&gt; means IP99 holds no current risk evidence for this address, not that the address is safe. &lt;code&gt;risk.score&lt;/code&gt; is the freshness and strength of verifiable evidence, not a probability that someone is an attacker. &lt;code&gt;geo&lt;/code&gt; placed the block in Zhengzhou; for anycast and carrier ranges, city-level geo is often a point of presence, not a building. An empty &lt;code&gt;district&lt;/code&gt; is a missing place name, not a zero. The same response arrived with &lt;code&gt;Access-Control-Allow-Origin: *&lt;/code&gt;, &lt;code&gt;x-daily-limit: 500&lt;/code&gt;, &lt;code&gt;x-ratelimit-limit: 60&lt;/code&gt;, and &lt;code&gt;x-plan: anon&lt;/code&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  Look up any other address the same way
&lt;/h2&gt;

&lt;p&gt;Once you know the egress, compare it with a well-known resolver, a customer address, or a previous snapshot:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/8.8.8.8
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Verified the same day (&lt;code&gt;computed_at&lt;/code&gt; &lt;code&gt;2026-11-15T13:14:23Z&lt;/code&gt;):&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="nl"&gt;"ip"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"8.8.8.8"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"computed_at"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"2026-11-15T13:14:23Z"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"evidence_state"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"geo"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"continent"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"NA"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"country"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"US"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"region"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"California"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"city"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Mountain View"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"tz"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"America/Los_Angeles"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"network"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"asn"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;15169&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"usage_type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"IDC"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"risk"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"score"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;0&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"level"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"signals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"hosting"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Same shape, same reading rules. &lt;code&gt;hosting&lt;/code&gt; is here because the usage type is &lt;code&gt;IDC&lt;/code&gt;, not because Google Public DNS is a proxy. The &lt;a href="https://ip99.com/api?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;API&lt;/a&gt; treats the address as a required path parameter when you pass one: private and reserved ranges return &lt;code&gt;400 reserved_ip&lt;/code&gt;; a malformed value returns &lt;code&gt;400 invalid_ip&lt;/code&gt;. Free &lt;code&gt;/v1/ip&lt;/code&gt; answers for now only; passing &lt;code&gt;?at=&lt;/code&gt; is &lt;code&gt;400 at_unsupported&lt;/code&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  Private addresses have no public profile
&lt;/h2&gt;

&lt;p&gt;If you paste the address from &lt;code&gt;ip addr&lt;/code&gt; or &lt;code&gt;ipconfig&lt;/code&gt; into a public lookup, you often get an error, not a "clean" score:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/192.168.1.1
curl https://ip99.com/v1/ip/10.0.0.1
curl https://ip99.com/v1/ip/127.0.0.1
curl https://ip99.com/v1/ip/169.254.1.1
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Verified 15 November 2026 UTC, all four returned HTTP 400:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="nl"&gt;"error"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="nl"&gt;"code"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="s2"&gt;"reserved_ip"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="nl"&gt;"message"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="s2"&gt;"private or reserved address has no public risk profile"&lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;&lt;code&gt;::1&lt;/code&gt; does the same. Official docs: branch on &lt;code&gt;error.code&lt;/code&gt;, never on &lt;code&gt;error.message&lt;/code&gt; — the message is prose for humans and its wording may change. These ranges are not on the public internet, so there is no public risk profile to return. The useful move is the no-path call: ask IP99 which address it actually received.&lt;/p&gt;

&lt;h2&gt;
  
  
  The exit is not always "you"
&lt;/h2&gt;

&lt;p&gt;A few cases change the address, or change what it means, without anything being "wrong":&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;A VPN or proxy replaces the egress. While one is active, the lookup reports the exit's network and country, not your access line. That is usually why two "what is my IP" tabs disagree after you toggle a client.&lt;/li&gt;
&lt;li&gt;Most home and mobile connections get an address by DHCP. A router reboot, a lease expiry, or a carrier re-addressing can move it. Query at decision time; do not cache a verdict overnight and treat it as still true.&lt;/li&gt;
&lt;li&gt;IPv4 and IPv6 can disagree. A dual-stack connection has both, and a site may see whichever its DNS prefers. If two services report different addresses, check which protocol each one used.&lt;/li&gt;
&lt;li&gt;Shared egress is normal. Carrier-grade NAT and some apartment or guest networks put many subscribers behind one public address (&lt;code&gt;usage_type&lt;/code&gt; is often &lt;code&gt;GTW&lt;/code&gt;). The address identifies a connection, not a person — official guidance is to step up (captcha, second factor, review), not to hard-ban a building.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;IP geolocation is a separate database from the address itself. It cannot name a person, an account, or a street number. City-level accuracy is the realistic case for consumer addresses; a result that points at an ISP regional office or a CDN PoP is expected, not a fault. See &lt;a href="https://ip99.com/en/ip-location?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;how IP location works&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why bother looking at your own exit
&lt;/h2&gt;

&lt;p&gt;The same call is useful before you debug someone else's:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;A CI runner, cloud function, or shared office gateway often sits on &lt;code&gt;IDC&lt;/code&gt; / &lt;code&gt;hosting&lt;/code&gt; space. If a vendor flags "datacenter," the no-path lookup tells you whether that is actually your pipeline's egress.&lt;/li&gt;
&lt;li&gt;Before you assume a geo or risk decision is a false positive, confirm the address the remote side received — especially if a VPN, IPv6, or a corporate proxy is in the path.&lt;/li&gt;
&lt;li&gt;In a browser, &lt;code&gt;GET /v1/ip&lt;/code&gt; and &lt;code&gt;GET /v1/ip/{ip}&lt;/code&gt; send &lt;code&gt;Access-Control-Allow-Origin: *&lt;/code&gt; and answer &lt;code&gt;OPTIONS&lt;/code&gt; preflight, so a page can call them directly. The anonymous allowance then counts against the visitor's own IP, not your origin server.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Anonymous quota is &lt;strong&gt;500 lookups/day&lt;/strong&gt; per client IP (60/min); a free logged-in key raises the daily cap to &lt;strong&gt;5,000/day&lt;/strong&gt;. Check remaining with &lt;code&gt;GET https://ip99.com/v1/me&lt;/code&gt; or the &lt;code&gt;X-Daily-*&lt;/code&gt; headers. OpenAPI: &lt;a href="https://ip99.com/openapi.json?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com/openapi.json&lt;/a&gt;. Pricing: &lt;a href="https://ip99.com/en/pricing?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com/en/pricing&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;Do &lt;strong&gt;not&lt;/strong&gt; treat IP99 scores as identical to other ThreatHunter datasets — this article only claims what free &lt;code&gt;/v1/ip&lt;/code&gt; and the cited product pages document.&lt;/p&gt;

&lt;h2&gt;
  
  
  Try it
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip
curl https://ip99.com/v1/ip/8.8.8.8
curl https://ip99.com/v1/ip/192.168.1.1
curl https://ip99.com/v1/me
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;IP99 is a ThreatHunter product. The address on your machine is not always the address the internet sees; start with the no-path lookup, and read &lt;code&gt;evidence_state&lt;/code&gt; before you treat a 0 as "safe." Start at &lt;a href="https://ip99.com/?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com&lt;/a&gt;; the field contract is on &lt;a href="https://ip99.com/en/what-is-my-ip?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;What is my IP address?&lt;/a&gt; and &lt;a href="https://ip99.com/api?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com/api&lt;/a&gt;.&lt;/p&gt;

</description>
      <category>api</category>
      <category>security</category>
      <category>networking</category>
      <category>cybersecurity</category>
    </item>
    <item>
      <title>An IP Risk Score Is Not Attacker Probability: Reading risk.score on IP99</title>
      <dc:creator>IP99</dc:creator>
      <pubDate>Sat, 14 Nov 2026 13:15:41 +0000</pubDate>
      <link>https://dev.to/ip99/an-ip-risk-score-is-not-attacker-probability-reading-riskscore-on-ip99-400p</link>
      <guid>https://dev.to/ip99/an-ip-risk-score-is-not-attacker-probability-reading-riskscore-on-ip99-400p</guid>
      <description>&lt;p&gt;Teams often treat an IP risk score as a fraud probability: 88 means "an 88% chance this user is an attacker," and 0 means "safe." That reading is the most common way an IP feed gets misused. On &lt;a href="https://ip99.com?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP99&lt;/a&gt; by ThreatHunter, &lt;code&gt;risk.score&lt;/code&gt; measures the &lt;strong&gt;freshness and strength of verifiable evidence&lt;/strong&gt; that an address is currently in risk infrastructure (proxy, VPN, dial-up pool, cloud phone, and similar signals), and the number decays as that evidence ages. It is not the probability that the person behind the address intends fraud.&lt;/p&gt;

&lt;p&gt;This walkthrough stays on the free, no-key profile &lt;code&gt;GET /v1/ip/{ip}&lt;/code&gt;. JSON samples were verified on 14 November 2026 UTC. Notes: &lt;a href="https://ip99.com?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP99&lt;/a&gt;, &lt;a href="https://ip99.com/api?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;API&lt;/a&gt;, &lt;a href="https://ip99.com/en/ip-risk-score?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP risk score&lt;/a&gt;, &lt;a href="https://ip99.com/en/ip-risk-evidence-freshness?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;evidence freshness&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  What the free lookup returns
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;code&gt;risk.score&lt;/code&gt; — 0–100, freshness and strength of verifiable risk evidence, decaying with its age.&lt;/li&gt;
&lt;li&gt;
&lt;code&gt;risk.level&lt;/code&gt; — &lt;code&gt;none&lt;/code&gt; (0), &lt;code&gt;low&lt;/code&gt; (1–29), &lt;code&gt;medium&lt;/code&gt; (30–69), &lt;code&gt;high&lt;/code&gt; (70–100).&lt;/li&gt;
&lt;li&gt;
&lt;code&gt;evidence_state&lt;/code&gt; — &lt;code&gt;active&lt;/code&gt; (fresh evidence that still scores), &lt;code&gt;stale&lt;/code&gt; (seen before, evidence expired, score 0, still a weak prior), &lt;code&gt;none&lt;/code&gt; (never seen).&lt;/li&gt;
&lt;li&gt;
&lt;code&gt;risk.signals&lt;/code&gt; — matched slugs only, alphabetical; the key is absent when nothing matched. &lt;code&gt;hosting&lt;/code&gt; and &lt;code&gt;mobile&lt;/code&gt; are network attributes, not risk verdicts — official guidance is not to block on those slugs alone.&lt;/li&gt;
&lt;li&gt;
&lt;code&gt;computed_at&lt;/code&gt; — when this verdict was computed.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Official docs are explicit: two addresses with the same score can carry very different fraud risk depending on your own signals. The score tells you what the &lt;strong&gt;network&lt;/strong&gt; has been seen doing, not what the user intends. Branch on &lt;code&gt;evidence_state&lt;/code&gt; &lt;strong&gt;before&lt;/strong&gt; the number.&lt;/p&gt;

&lt;h2&gt;
  
  
  The same 0, two different meanings
&lt;/h2&gt;

&lt;p&gt;Start with a well-known resolver:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/8.8.8.8
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Verified 14 November 2026 UTC (&lt;code&gt;computed_at&lt;/code&gt; &lt;code&gt;2026-11-14T13:06:08Z&lt;/code&gt;):&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="nl"&gt;"ip"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"8.8.8.8"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"computed_at"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"2026-11-14T13:06:08Z"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"evidence_state"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"geo"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"continent"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"NA"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"country"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"US"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"region"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"California"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"city"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Mountain View"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"tz"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"America/Los_Angeles"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"network"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"asn"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;15169&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"usage_type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"IDC"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"risk"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"score"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;0&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"level"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"signals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"hosting"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Score 0 and &lt;code&gt;evidence_state: "none"&lt;/code&gt; mean IP99 holds no current risk evidence for this address. &lt;code&gt;hosting&lt;/code&gt; is here because the usage type is &lt;code&gt;IDC&lt;/code&gt;, not because Google Public DNS is a proxy. Marking 8.8.8.8 "safe" because the score is 0, or blocking it because &lt;code&gt;hosting&lt;/code&gt; is present, both misread the fields.&lt;/p&gt;

&lt;p&gt;Now look up an address that used to score high:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/78.141.167.222
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Same day (&lt;code&gt;computed_at&lt;/code&gt; &lt;code&gt;2026-11-14T13:06:17Z&lt;/code&gt;):&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="nl"&gt;"ip"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"78.141.167.222"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"computed_at"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"2026-11-14T13:06:17Z"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"evidence_state"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"stale"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"geo"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"continent"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"EU"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"country"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"LU"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"region"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Luxembourg"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"city"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Luxembourg"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"tz"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Europe/Luxembourg"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"network"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"asn"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;6661&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"usage_type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"DYN"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"risk"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"score"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;0&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"level"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"signals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"proxy"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;The score is also 0. The meaning is not. &lt;code&gt;evidence_state: "stale"&lt;/code&gt; plus &lt;code&gt;signals: ["proxy"]&lt;/code&gt; means this address was observed as a proxy exit, and that evidence has aged past the decay window. On 9 October 2026 UTC, a lookup of the same address returned &lt;code&gt;evidence_state: "active"&lt;/code&gt;, score &lt;code&gt;97&lt;/code&gt;, level &lt;code&gt;high&lt;/code&gt;, and &lt;code&gt;proxy&lt;/code&gt; on residential (&lt;code&gt;DYN&lt;/code&gt;) space. Today the number is 0 because the judgement expired — not because the product "cleared" the address.&lt;/p&gt;

&lt;p&gt;That is the official distinction on the &lt;a href="https://ip99.com/en/ip-risk-score?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP risk score&lt;/a&gt; guide: 0 + &lt;code&gt;none&lt;/code&gt; is "unknown"; 0 + &lt;code&gt;stale&lt;/code&gt; is "was, and probably no longer is." Neither is a clean bill of health. If your rule is &lt;code&gt;if score == 0: allow&lt;/code&gt;, both of these addresses take the same path, and you have thrown away the only field that distinguishes them.&lt;/p&gt;

&lt;h2&gt;
  
  
  The score can decay while it is still active
&lt;/h2&gt;

&lt;p&gt;A documented sample on IP99's &lt;a href="https://ip99.com/en/ip-risk-evidence-freshness?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;evidence freshness&lt;/a&gt; page was captured on 13 September 2026, when &lt;code&gt;122.99.101.50&lt;/code&gt; was being sold as a residential exit: &lt;code&gt;evidence_state: "active"&lt;/code&gt;, score &lt;code&gt;78&lt;/code&gt;, level &lt;code&gt;high&lt;/code&gt;, &lt;code&gt;signals: ["proxy"]&lt;/code&gt;, &lt;code&gt;usage_type: "DYN"&lt;/code&gt;, ASN 17471, &lt;code&gt;computed_at&lt;/code&gt; &lt;code&gt;2026-09-13T15:41:18Z&lt;/code&gt;.&lt;/p&gt;

&lt;p&gt;The same address today (&lt;code&gt;computed_at&lt;/code&gt; &lt;code&gt;2026-11-14T13:06:22Z&lt;/code&gt;) is still &lt;code&gt;active&lt;/code&gt; and still &lt;code&gt;proxy&lt;/code&gt; on &lt;code&gt;DYN&lt;/code&gt; space in Dhaka, ASN 17471, but the score is &lt;code&gt;16&lt;/code&gt; and the level is &lt;code&gt;low&lt;/code&gt;. The snapshot also carried &lt;code&gt;degraded: true&lt;/code&gt;, meaning some data sources were unavailable when the verdict was computed; the fields above are still the live free-profile answer.&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/122.99.101.50
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;That is the score working as designed: still &lt;code&gt;active&lt;/code&gt;, but weaker than in September. Caching "this IP is high-risk" from September, or even from yesterday, turns a true statement into a false one on a timer you cannot see.&lt;/p&gt;

&lt;h2&gt;
  
  
  A high score is not a ban reason by itself
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/178.20.30.146
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Verified the same day (&lt;code&gt;computed_at&lt;/code&gt; &lt;code&gt;2026-11-14T13:06:20Z&lt;/code&gt;; this snapshot also had &lt;code&gt;degraded: true&lt;/code&gt;):&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="nl"&gt;"ip"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"178.20.30.146"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"computed_at"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"2026-11-14T13:06:20Z"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"evidence_state"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"active"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"geo"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"continent"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"EU"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"country"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"FR"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"region"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Île-de-France"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"city"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Paris"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"tz"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Europe/Paris"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"network"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"asn"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;46475&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"usage_type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"IDC"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"risk"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"score"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;88&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"level"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"high"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"signals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"hosting"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"osint"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"proxy"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;88 is not "an 88% chance this session is fraud." It means the currently verifiable evidence that this datacenter address is a proxy exit (and is listed in OSINT sources) is strong and still fresh. Official guidance is to &lt;strong&gt;step up&lt;/strong&gt; — a second factor, a delay, or manual review — not to hard-ban, especially where carrier-grade NAT and residential pools share exits with real customers. &lt;code&gt;osint&lt;/code&gt; is a list hit, not a probability that the operator is an attacker; &lt;code&gt;hosting&lt;/code&gt; is the usage-type attribute again.&lt;/p&gt;

&lt;h2&gt;
  
  
  How to branch on it
&lt;/h2&gt;

&lt;p&gt;The official shape:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight python"&gt;&lt;code&gt;&lt;span class="k"&gt;if&lt;/span&gt; &lt;span class="n"&gt;evidence_state&lt;/span&gt; &lt;span class="o"&gt;==&lt;/span&gt; &lt;span class="s2"&gt;"active"&lt;/span&gt; &lt;span class="ow"&gt;and&lt;/span&gt; &lt;span class="n"&gt;risk&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="n"&gt;level&lt;/span&gt; &lt;span class="k"&gt;in&lt;/span&gt; &lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="s2"&gt;"high"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"medium"&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
    &lt;span class="nf"&gt;step_up&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;          &lt;span class="c1"&gt;# second factor, delay, manual review — not a ban&lt;/span&gt;
&lt;span class="k"&gt;elif&lt;/span&gt; &lt;span class="n"&gt;evidence_state&lt;/span&gt; &lt;span class="o"&gt;==&lt;/span&gt; &lt;span class="s2"&gt;"stale"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
    &lt;span class="nf"&gt;weak_prior&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;       &lt;span class="c1"&gt;# combine with device, velocity, account age&lt;/span&gt;
&lt;span class="k"&gt;else&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;                  &lt;span class="c1"&gt;# "none"&lt;/span&gt;
    &lt;span class="nf"&gt;no_information&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;   &lt;span class="c1"&gt;# NOT "safe": fall back to your other signals&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Two habits keep this honest: query at &lt;strong&gt;decision time&lt;/strong&gt; (reuse one response only for the seconds of a single decision), and store &lt;code&gt;computed_at&lt;/code&gt; plus &lt;code&gt;evidence_state&lt;/code&gt; with the decision so a disputed block can be traced to when it was observed. Free &lt;code&gt;/v1/ip&lt;/code&gt; answers for now only and rejects &lt;code&gt;?at=&lt;/code&gt; with &lt;code&gt;400 at_unsupported&lt;/code&gt;. Historical replay is a Pulse feature (&lt;code&gt;?at=&lt;/code&gt; on &lt;code&gt;/pulse/v1/ip/{ip}&lt;/code&gt;, up to 30 days). Without a professional key, Pulse returns &lt;code&gt;403 professional_required&lt;/code&gt;.&lt;/p&gt;

&lt;p&gt;Generic triage (not a contract): score &lt;code&gt;0&lt;/code&gt; → do not stamp "clean"; &lt;code&gt;1–29&lt;/code&gt; → observe; &lt;code&gt;30–69&lt;/code&gt; → captcha / throttle / second factor; &lt;code&gt;70–100&lt;/code&gt; → block only when business cost allows, and not at building or prefix scale. Official order: &lt;strong&gt;evidence_state → score/level → signal type&lt;/strong&gt;, then &lt;code&gt;usage_type&lt;/code&gt; on the free profile (&lt;code&gt;share_tier&lt;/code&gt; on Pulse). If &lt;code&gt;hijacked_proxy&lt;/code&gt; is present, challenge the session; do not ban the account — the address holder is the victim.&lt;/p&gt;

&lt;p&gt;Anonymous quota is &lt;strong&gt;500 lookups/day&lt;/strong&gt; per client IP (60/min); a free logged-in key raises the daily cap to &lt;strong&gt;5,000/day&lt;/strong&gt;. Check remaining with &lt;code&gt;GET https://ip99.com/v1/me&lt;/code&gt; or the &lt;code&gt;X-Daily-*&lt;/code&gt; headers. OpenAPI: &lt;a href="https://ip99.com/openapi.json?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com/openapi.json&lt;/a&gt;. Pricing: &lt;a href="https://ip99.com/en/pricing?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com/en/pricing&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;Do &lt;strong&gt;not&lt;/strong&gt; treat IP99 scores as identical to other ThreatHunter datasets — this article only claims what free &lt;code&gt;/v1/ip&lt;/code&gt; and the cited product pages document.&lt;/p&gt;

&lt;h2&gt;
  
  
  Try it
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/8.8.8.8
curl https://ip99.com/v1/ip/78.141.167.222
curl https://ip99.com/v1/ip/122.99.101.50
curl https://ip99.com/v1/ip/178.20.30.146
curl https://ip99.com/v1/me
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;IP99 is a ThreatHunter product. Read &lt;code&gt;evidence_state&lt;/code&gt; before you treat a 0 as "safe" or an 88 as a probability. Start at &lt;a href="https://ip99.com?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com&lt;/a&gt;; the field contract is on &lt;a href="https://ip99.com/en/ip-risk-score?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;What an IP risk score means&lt;/a&gt;, &lt;a href="https://ip99.com/en/ip-risk-evidence-freshness?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;evidence freshness&lt;/a&gt;, and &lt;a href="https://ip99.com/api?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com/api&lt;/a&gt;.&lt;/p&gt;

</description>
      <category>security</category>
      <category>api</category>
      <category>cybersecurity</category>
      <category>networking</category>
    </item>
    <item>
      <title>Your IPv4 and IPv6 Exits Can Disagree: Looking Up IPv6 on IP99</title>
      <dc:creator>IP99</dc:creator>
      <pubDate>Fri, 13 Nov 2026 13:13:06 +0000</pubDate>
      <link>https://dev.to/ip99/your-ipv4-and-ipv6-exits-can-disagree-looking-up-ipv6-on-ip99-54pb</link>
      <guid>https://dev.to/ip99/your-ipv4-and-ipv6-exits-can-disagree-looking-up-ipv6-on-ip99-54pb</guid>
      <description>&lt;p&gt;A dual-stack client has two public identities, not one. If your risk check only looks up the IPv4, you have answered half the question: the IPv6 on the same line can show a different city, and a site may have used a different protocol. On IP99 by ThreatHunter, free, no-key &lt;code&gt;GET /v1/ip/{ip}&lt;/code&gt; accepts both IPv4 and IPv6. &lt;code&gt;GET /v1/ip&lt;/code&gt; (no address in the path) returns the caller's current public exit — whichever protocol your client used to reach the API.&lt;/p&gt;

&lt;p&gt;This walkthrough stays on that free profile. JSON samples were verified on 13 November 2026 UTC. Notes: &lt;a href="https://ip99.com?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP99&lt;/a&gt;, &lt;a href="https://ip99.com/api?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;API&lt;/a&gt;, &lt;a href="https://ip99.com/en/what-is-my-ip?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;What is my IP&lt;/a&gt;, &lt;a href="https://ip99.com/en/ip-location?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP location&lt;/a&gt;, &lt;a href="https://ip99.com/en/ip-risk-score?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP risk score&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  What the free lookup returns
&lt;/h2&gt;

&lt;p&gt;&lt;code&gt;risk.score&lt;/code&gt; is 0–100 for &lt;strong&gt;freshness and strength of verifiable risk evidence&lt;/strong&gt;, not attacker probability. &lt;code&gt;risk.level&lt;/code&gt; follows the score (&lt;code&gt;0&lt;/code&gt; → &lt;code&gt;none&lt;/code&gt;, &lt;code&gt;1–29&lt;/code&gt; → &lt;code&gt;low&lt;/code&gt;, &lt;code&gt;30–69&lt;/code&gt; → &lt;code&gt;medium&lt;/code&gt;, &lt;code&gt;70–100&lt;/code&gt; → &lt;code&gt;high&lt;/code&gt;). Signals appear only when they hit. Branch on &lt;code&gt;evidence_state&lt;/code&gt; (&lt;code&gt;active&lt;/code&gt; / &lt;code&gt;stale&lt;/code&gt; / &lt;code&gt;none&lt;/code&gt;) &lt;strong&gt;before&lt;/strong&gt; the score: &lt;code&gt;0&lt;/code&gt; + &lt;code&gt;none&lt;/code&gt; is a coverage gap, not a clean bill of health. &lt;code&gt;hosting&lt;/code&gt; and &lt;code&gt;mobile&lt;/code&gt; are network attributes, not risk verdicts — official guidance is not to block on those slugs alone.&lt;/p&gt;

&lt;p&gt;The path parameter is a &lt;strong&gt;bare address&lt;/strong&gt;: an IPv4 or IPv6 literal, no port, no zone id.&lt;/p&gt;

&lt;h2&gt;
  
  
  Same resolver family, two protocols
&lt;/h2&gt;

&lt;p&gt;Google Public DNS agrees with itself in this snapshot. IPv4 and IPv6 land in the same city, the same ASN, and the same empty evidence set:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl -g https://ip99.com/v1/ip/8.8.8.8
curl -g https://ip99.com/v1/ip/2001:4860:4860::8888
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Verified 13 November 2026 UTC (&lt;code&gt;2001:4860:4860::8888&lt;/code&gt;, &lt;code&gt;computed_at&lt;/code&gt; &lt;code&gt;2026-11-13T13:06:06Z&lt;/code&gt;):&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="nl"&gt;"ip"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"2001:4860:4860::8888"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"computed_at"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"2026-11-13T13:06:06Z"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"evidence_state"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"geo"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"continent"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"NA"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"country"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"US"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"region"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"California"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"city"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Mountain View"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"tz"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"America/Los_Angeles"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"network"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"asn"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;15169&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"usage_type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"IDC"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"risk"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"score"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;0&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"level"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"signals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"hosting"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;&lt;code&gt;8.8.8.8&lt;/code&gt; in the same minute is the same shape: Mountain View, ASN 15169, &lt;code&gt;usage_type: IDC&lt;/code&gt;, &lt;code&gt;hosting&lt;/code&gt;, score &lt;code&gt;0&lt;/code&gt;, &lt;code&gt;evidence_state: none&lt;/code&gt;. Agreement is a possible outcome, not a rule. &lt;code&gt;hosting&lt;/code&gt; here is infrastructure, not a reason to refuse the resolver.&lt;/p&gt;

&lt;p&gt;Cloudflare's public resolver does &lt;strong&gt;not&lt;/strong&gt; agree on geography in the same snapshot:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl -g https://ip99.com/v1/ip/1.1.1.1
curl -g https://ip99.com/v1/ip/2606:4700:4700::1111
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Both addresses are ASN 13335, &lt;code&gt;usage_type: IDC&lt;/code&gt;, &lt;code&gt;hosting&lt;/code&gt;, score &lt;code&gt;0&lt;/code&gt;, &lt;code&gt;evidence_state: none&lt;/code&gt;. The IPv4 (&lt;code&gt;computed_at&lt;/code&gt; &lt;code&gt;2026-11-13T13:06:12Z&lt;/code&gt;) is placed in Brisbane, Queensland, AU (&lt;code&gt;Australia/Brisbane&lt;/code&gt;). The IPv6 (&lt;code&gt;2026-11-13T13:06:10Z&lt;/code&gt;) is placed in San Francisco, California, US (&lt;code&gt;America/Los_Angeles&lt;/code&gt;). That is two anycast exits, not a person moving house.&lt;/p&gt;

&lt;p&gt;IP99's &lt;a href="https://ip99.com/en/what-is-my-ip?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;What is my IP&lt;/a&gt; page says this in product language: a dual-stack connection has a public IPv4 address and an IPv6 prefix, and a site may see whichever protocol its DNS prefers. If two services disagree about "your" address, check which protocol each one used. Geolocation places the &lt;strong&gt;block&lt;/strong&gt;, not the building; city-level is the realistic accuracy for consumer addresses.&lt;/p&gt;

&lt;p&gt;Treat the IPv4 and the IPv6 as &lt;strong&gt;two records&lt;/strong&gt;. Do not copy &lt;code&gt;geo&lt;/code&gt;, &lt;code&gt;network&lt;/code&gt;, or &lt;code&gt;risk&lt;/code&gt; from one onto the other.&lt;/p&gt;

&lt;h2&gt;
  
  
  How the path handles IPv6
&lt;/h2&gt;

&lt;p&gt;A compressed literal and a fully expanded one resolve to the same profile. In this snapshot, &lt;code&gt;2001:4860:4860:0:0:0:0:8888&lt;/code&gt; is returned as &lt;code&gt;ip: "2001:4860:4860::8888"&lt;/code&gt;. The unencoded literal is enough for &lt;code&gt;curl -g&lt;/code&gt; (URL-encoding colons also works).&lt;/p&gt;

&lt;p&gt;An IPv4-mapped address is answered as IPv4. &lt;code&gt;GET /v1/ip/::ffff:8.8.8.8&lt;/code&gt; returned the &lt;code&gt;8.8.8.8&lt;/code&gt; profile (&lt;code&gt;ip&lt;/code&gt; is &lt;code&gt;8.8.8.8&lt;/code&gt;, not the mapped form). Store the request string and the response &lt;code&gt;ip&lt;/code&gt; separately.&lt;/p&gt;

&lt;p&gt;Private, unique-local, loopback, and link-local addresses return HTTP 400 &lt;code&gt;reserved_ip&lt;/code&gt; ("private or reserved address has no public risk profile"). Verified here: &lt;code&gt;192.168.1.1&lt;/code&gt;, &lt;code&gt;::1&lt;/code&gt;, &lt;code&gt;fc00::1&lt;/code&gt;, and &lt;code&gt;fe80::1&lt;/code&gt;. Official docs: those ranges are not on the public internet, so they have no public profile.&lt;/p&gt;

&lt;p&gt;The documentation prefix is different. &lt;code&gt;2001:db8::1&lt;/code&gt; returned HTTP 200, not 400:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="nl"&gt;"ip"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"2001:db8::1"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"computed_at"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"2026-11-13T13:06:22Z"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"evidence_state"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"geo"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"continent"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;""&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"country"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;""&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"region"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"reserve"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"city"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"reserve"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"tz"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;""&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"network"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"usage_type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"BOGON"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"risk"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"score"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;0&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"level"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;No ASN, empty country, &lt;code&gt;usage_type: BOGON&lt;/code&gt;, and no &lt;code&gt;risk.signals&lt;/code&gt; key. That is not a customer session; a missing &lt;code&gt;asn&lt;/code&gt; is unknown, not zero.&lt;/p&gt;

&lt;h2&gt;
  
  
  Look up the exit you actually have
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;The &lt;code&gt;ip&lt;/code&gt; field is the public address the API observed on that connection — including a VPN or proxy exit if one is active, in which case the country is the exit's, not your line's. Official copy also notes that many home connections share one public IPv4, or one IPv6 prefix, behind the router, and that carrier-grade NAT means the address is not a person. To see both protocols, request once over IPv4 and once over IPv6. Re-query at decision time and store &lt;code&gt;computed_at&lt;/code&gt;; scores decay, and the pair can change independently.&lt;/p&gt;

&lt;h2&gt;
  
  
  Don't ban a single IPv6 and call it done
&lt;/h2&gt;

&lt;p&gt;A household IPv6 allocation is typically a prefix (commonly a &lt;code&gt;/56&lt;/code&gt; or &lt;code&gt;/64&lt;/code&gt;), and devices may rotate the interface identifier (privacy extensions, RFC 8981). A hard ban of one IPv6 address is a narrow lever: it can miss the next address from the same prefix, or punish every device behind that prefix if you widen the ban blindly.&lt;/p&gt;

&lt;p&gt;Free &lt;code&gt;/v1/ip&lt;/code&gt; answers &lt;strong&gt;one address&lt;/strong&gt;. It does not return &lt;code&gt;share_tier&lt;/code&gt;, &lt;code&gt;evidence[]&lt;/code&gt;, or a prefix-aggregation field. Those details, including Pulse-only &lt;code&gt;ipv6_range&lt;/code&gt; on &lt;code&gt;evidence[].risk_type&lt;/code&gt; (not a &lt;code&gt;risk.signals&lt;/code&gt; slug), live on &lt;code&gt;GET /pulse/v1/ip/{ip}&lt;/code&gt; (professional). Historical replay with &lt;code&gt;?at=&lt;/code&gt; is a Pulse feature. Official docs list &lt;code&gt;history_no_ipv6&lt;/code&gt; when an access point has not enabled IPv6 history (the history index is IPv4 until that is turned on). Free &lt;code&gt;/v1/ip&lt;/code&gt; rejects &lt;code&gt;?at=&lt;/code&gt; with &lt;code&gt;400 at_unsupported&lt;/code&gt;. Without a professional key, Pulse returns &lt;code&gt;403 professional_required&lt;/code&gt;.&lt;/p&gt;

&lt;p&gt;Generic triage (not a contract): score &lt;code&gt;0&lt;/code&gt; → do not stamp "clean"; &lt;code&gt;1–29&lt;/code&gt; → observe; &lt;code&gt;30–69&lt;/code&gt; → captcha / throttle / second factor; &lt;code&gt;70–100&lt;/code&gt; → block only when business cost allows, and not at building or prefix scale. Official order: &lt;strong&gt;evidence_state → score/level → signal type&lt;/strong&gt;, then &lt;code&gt;usage_type&lt;/code&gt; on the free profile (&lt;code&gt;share_tier&lt;/code&gt; on Pulse). If &lt;code&gt;hijacked_proxy&lt;/code&gt; is present, challenge the session; do not ban the account — the address holder is the victim.&lt;/p&gt;

&lt;p&gt;A practical branch for dual-stack traffic:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;Look up the IPv4 you observed &lt;strong&gt;and&lt;/strong&gt; the IPv6 you observed. If you only have a self-check, call &lt;code&gt;GET /v1/ip&lt;/code&gt; once per protocol.&lt;/li&gt;
&lt;li&gt;Keep two records. Never copy geo or score across families.&lt;/li&gt;
&lt;li&gt;Read &lt;code&gt;evidence_state&lt;/code&gt;, then &lt;code&gt;risk.signals&lt;/code&gt;, then &lt;code&gt;network.usage_type&lt;/code&gt; and &lt;code&gt;network.asn&lt;/code&gt;.&lt;/li&gt;
&lt;li&gt;On &lt;code&gt;MOB&lt;/code&gt; or &lt;code&gt;GTW&lt;/code&gt;, assume a wide NAT and prefer a step-up over a hard IP ban. Do not hard-ban a single consumer IPv6, and do not allowlist "it is IPv6."&lt;/li&gt;
&lt;li&gt;Never mark score &lt;code&gt;0&lt;/code&gt; + &lt;code&gt;evidence_state: none&lt;/code&gt; as "trusted clean."&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;Anonymous quota is &lt;strong&gt;500 lookups/day&lt;/strong&gt; per client IP (60/min); a free logged-in key raises the daily cap to &lt;strong&gt;5,000/day&lt;/strong&gt;. Check remaining with &lt;code&gt;GET https://ip99.com/v1/me&lt;/code&gt; or the &lt;code&gt;X-Daily-*&lt;/code&gt; headers. OpenAPI: &lt;a href="https://ip99.com/openapi.json?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com/openapi.json&lt;/a&gt;. Pricing: &lt;a href="https://ip99.com/en/pricing?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com/en/pricing&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;Do &lt;strong&gt;not&lt;/strong&gt; treat IP99 scores as identical to other ThreatHunter datasets — this article only claims what free &lt;code&gt;/v1/ip&lt;/code&gt; and the cited product pages document.&lt;/p&gt;

&lt;h2&gt;
  
  
  Try it
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip
curl -g https://ip99.com/v1/ip/8.8.8.8
curl -g https://ip99.com/v1/ip/2001:4860:4860::8888
curl -g https://ip99.com/v1/ip/1.1.1.1
curl -g https://ip99.com/v1/ip/2606:4700:4700::1111
curl -g https://ip99.com/v1/ip/2001:db8::1
curl https://ip99.com/v1/me
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;IP99 is a ThreatHunter product. Use the free lookup on &lt;strong&gt;both&lt;/strong&gt; families before you decide that an egress, a resolver, or a customer session is "the" address. Start at &lt;a href="https://ip99.com?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com&lt;/a&gt;; read &lt;a href="https://ip99.com/en/what-is-my-ip?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;What is my IP&lt;/a&gt;, &lt;a href="https://ip99.com/en/ip-location?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP location&lt;/a&gt;, and &lt;a href="https://ip99.com/api?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com/api&lt;/a&gt;.&lt;/p&gt;

</description>
      <category>security</category>
      <category>api</category>
      <category>ipv6</category>
      <category>networking</category>
    </item>
    <item>
      <title>This Week's Proxy Exits: 10.0M New Addresses, Still Almost All Off Datacenter Space</title>
      <dc:creator>IP99</dc:creator>
      <pubDate>Thu, 12 Nov 2026 13:11:26 +0000</pubDate>
      <link>https://dev.to/ip99/this-weeks-proxy-exits-100m-new-addresses-still-almost-all-off-datacenter-space-1dgg</link>
      <guid>https://dev.to/ip99/this-weeks-proxy-exits-100m-new-addresses-still-almost-all-off-datacenter-space-1dgg</guid>
      <description>&lt;p&gt;Datacenter blocklists still look like the default proxy defence in a lot of fraud stacks. The addresses showing up in proxy supply this week were still mostly not in datacenters. The mix shift worth flagging is the datacenter slice itself: it eased from &lt;strong&gt;3.2%&lt;/strong&gt; in the last posted recap back to &lt;strong&gt;2.2%&lt;/strong&gt;, in line with the flat ~2% band of mid-to-late October.&lt;/p&gt;

&lt;p&gt;Between &lt;strong&gt;5–11 November 2026 (UTC)&lt;/strong&gt;, the &lt;a href="https://ip99.com?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP99&lt;/a&gt; risk feed captured &lt;strong&gt;10,026,346&lt;/strong&gt; proxy exit addresses it had not seen in the previous 7 days (&lt;strong&gt;−5%&lt;/strong&gt; vs the previous 7 days). Source page: &lt;a href="https://ip99.com/en/proxy-exits?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;ip99.com/en/proxy-exits&lt;/a&gt; (generated 12 November 2026, 12:36 UTC). The public page rolls daily over the last 7 complete UTC days; this is the window it reports today, continuing the series after the 4–10 November recap (the two windows overlap on 5–10 November).&lt;/p&gt;

&lt;p&gt;Against that last posted week (&lt;strong&gt;10,561,845&lt;/strong&gt; addresses), volume is also down about &lt;strong&gt;−5.1%&lt;/strong&gt;. The page's official &lt;strong&gt;−5%&lt;/strong&gt; compares 5–11 November with 29 October–4 November, not with the overlapping posted window. A large part of the step-down is compositional: &lt;strong&gt;11 November&lt;/strong&gt; was a lighter day (&lt;strong&gt;1,217,179&lt;/strong&gt; captured) and replaced &lt;strong&gt;4 November&lt;/strong&gt; (&lt;strong&gt;1,752,678&lt;/strong&gt;) as the window rolled forward. That describes what entered the 7-day mirror, not worldwide proxy supply.&lt;/p&gt;

&lt;h2&gt;
  
  
  What those addresses sit on
&lt;/h2&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Usage type&lt;/th&gt;
&lt;th&gt;Addresses&lt;/th&gt;
&lt;th&gt;Share&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;Residential broadband&lt;/td&gt;
&lt;td&gt;8,092,828&lt;/td&gt;
&lt;td&gt;80.7%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Mobile network&lt;/td&gt;
&lt;td&gt;1,227,350&lt;/td&gt;
&lt;td&gt;12.2%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Gateway / NAT egress&lt;/td&gt;
&lt;td&gt;405,698&lt;/td&gt;
&lt;td&gt;4.0%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Datacenter&lt;/td&gt;
&lt;td&gt;221,094&lt;/td&gt;
&lt;td&gt;2.2%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Education network&lt;/td&gt;
&lt;td&gt;50,409&lt;/td&gt;
&lt;td&gt;0.5%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Unclassified&lt;/td&gt;
&lt;td&gt;28,967&lt;/td&gt;
&lt;td&gt;0.3%&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;About &lt;strong&gt;62%&lt;/strong&gt; were rated high-risk at the moment of capture. The rest still carried a proxy-class signal at a lower risk level.&lt;/p&gt;

&lt;p&gt;IP99 counts an address as "captured" when it enters the feed with a proxy-class signal (&lt;code&gt;proxy&lt;/code&gt;, &lt;code&gt;VPN&lt;/code&gt;, dial-up pool, cloud phone, or hijacked shared proxy) and was &lt;strong&gt;not&lt;/strong&gt; in the live 7-day risk mirror. Addresses that go quiet and return can be counted again. The page describes what &lt;strong&gt;56&lt;/strong&gt; upstream collection sources delivered — not the total number of proxies in the world. An address is counted once per window. No address lists are published.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why 2.2% still is not "mostly datacenter"
&lt;/h2&gt;

&lt;p&gt;If your defence is mainly an ASN or hosting blocklist, you are still mostly looking past the problem. Residential broadband alone is &lt;strong&gt;80.7%&lt;/strong&gt; of this week's capture window. Mobile adds another &lt;strong&gt;12.2%&lt;/strong&gt;. Together that is &lt;strong&gt;93%&lt;/strong&gt; of newly seen proxy exits sitting in access networks where real customers also live.&lt;/p&gt;

&lt;p&gt;The datacenter row moved from &lt;strong&gt;3.2%&lt;/strong&gt; in the last posted recap (4–10 Nov) and &lt;strong&gt;3.6%&lt;/strong&gt; in the 29 Oct–4 Nov recap down to &lt;strong&gt;2.2%&lt;/strong&gt; here, back near the &lt;strong&gt;2.0%&lt;/strong&gt; of the 26 Oct–1 Nov window a fortnight earlier. The headline does not flip. A hosting-only blocklist still misses the majority of the supply-side traffic that shows up as residential or mobile exits.&lt;/p&gt;

&lt;p&gt;Usage type answers "what kind of network is this," not "is this address currently sold as an exit." Rotating residential proxies and dial-up pools are designed to sit in DYN / MOB space. Detecting them needs supply-side observation with an age attached, not a permanent network-type label. Product notes: &lt;a href="https://ip99.com/en/rotating-residential-proxies?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;rotating residential proxies&lt;/a&gt;, &lt;a href="https://ip99.com/en/detect-proxy-vpn?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;detect proxy / VPN&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  Which networks show up most
&lt;/h2&gt;

&lt;p&gt;Top ASNs in this week's capture window (by distinct address count):&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;ASN&lt;/th&gt;
&lt;th&gt;Addresses&lt;/th&gt;
&lt;th&gt;Share&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;AS4134&lt;/td&gt;
&lt;td&gt;946,807&lt;/td&gt;
&lt;td&gt;9.4%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;AS4837&lt;/td&gt;
&lt;td&gt;726,730&lt;/td&gt;
&lt;td&gt;7.2%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;AS9808&lt;/td&gt;
&lt;td&gt;356,305&lt;/td&gt;
&lt;td&gt;3.6%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;AS6057&lt;/td&gt;
&lt;td&gt;256,393&lt;/td&gt;
&lt;td&gt;2.6%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;AS36903&lt;/td&gt;
&lt;td&gt;199,394&lt;/td&gt;
&lt;td&gt;2.0%&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;AS4134, AS4837, and AS9808 remain the three largest contributors, as they have in recent recaps. The top five is unchanged from the 4–10 November recap, with AS6057 and AS36903 still fourth and fifth. AS55836, which held fifth place in the 29 Oct–4 Nov recap (2.2%), again sits eighth this week (1.6%). Large access networks appearing here are a clue about where proxy supply is being sold from — not a claim that every customer on those networks is a proxy. Ban a whole ASN and you take out legitimate home and mobile users. Prefer per-address evidence at decision time. ASN pages for deeper network mix: &lt;a href="https://ip99.com/en/asn/?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;ip99.com/en/asn/&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  Day-to-day shape
&lt;/h2&gt;

&lt;p&gt;Captured addresses (new to the prior 7-day mirror that day), with how many were high-risk at capture:&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Day (UTC)&lt;/th&gt;
&lt;th&gt;Captured&lt;/th&gt;
&lt;th&gt;High-risk at capture&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-05&lt;/td&gt;
&lt;td&gt;1,493,146&lt;/td&gt;
&lt;td&gt;946,402&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-06&lt;/td&gt;
&lt;td&gt;1,544,553&lt;/td&gt;
&lt;td&gt;943,591&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-07&lt;/td&gt;
&lt;td&gt;1,517,140&lt;/td&gt;
&lt;td&gt;919,244&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-08&lt;/td&gt;
&lt;td&gt;1,415,823&lt;/td&gt;
&lt;td&gt;897,828&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-09&lt;/td&gt;
&lt;td&gt;1,371,503&lt;/td&gt;
&lt;td&gt;883,005&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-10&lt;/td&gt;
&lt;td&gt;1,467,002&lt;/td&gt;
&lt;td&gt;940,244&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-11&lt;/td&gt;
&lt;td&gt;1,217,179&lt;/td&gt;
&lt;td&gt;695,544&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;Volume was down week-over-week (official &lt;strong&gt;−5%&lt;/strong&gt;), while upstream sources sat at &lt;strong&gt;56&lt;/strong&gt; (59 in the last posted recap). Daily capture stayed in a band roughly &lt;strong&gt;1.22–1.54M&lt;/strong&gt; addresses, with &lt;strong&gt;11 Nov&lt;/strong&gt; the lighter day and &lt;strong&gt;6 Nov&lt;/strong&gt; the heavier one in this window — still not a shape that makes a static list look "caught up."&lt;/p&gt;

&lt;h2&gt;
  
  
  How to use this without over-blocking
&lt;/h2&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Branch on evidence, not network type.&lt;/strong&gt; Residential is where your customers live; do not hard-block on DYN / MOB alone.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Step up, don't stop.&lt;/strong&gt; A fresh proxy-class signal on signup, a coupon claim, or a login from a new device justifies a second factor or a delay — not an immediate ban of a whole building's shared egress.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Read the age.&lt;/strong&gt; &lt;code&gt;evidence_state: "stale"&lt;/code&gt; means the address was in supply before and is not now; treat it as a weak prior. Re-query at decision time rather than caching a verdict for days. Score &lt;code&gt;0&lt;/code&gt; with &lt;code&gt;evidence_state: "none"&lt;/code&gt; is not a clean bill of health either — it means no current verifiable evidence is on file.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Keep the timestamp.&lt;/strong&gt; Store &lt;code&gt;computed_at&lt;/code&gt; with the decision so a disputed block can be traced back to when the observation was made.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;IP99 by ThreatHunter publishes the free weekly aggregate so you can see the mix without buying a feed. The score on a single lookup is the freshness and strength of verifiable risk evidence — not a probability that the address "is an attacker." Do not treat IP99 scores as identical to every other ThreatHunter dataset you may have seen elsewhere; this article only claims what the public proxy-exits page and free &lt;code&gt;/v1/ip&lt;/code&gt; document.&lt;/p&gt;

&lt;h2&gt;
  
  
  Check one yourself
&lt;/h2&gt;

&lt;p&gt;Free, no signup:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/8.8.8.8
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Look at &lt;code&gt;evidence_state&lt;/code&gt; (&lt;code&gt;active&lt;/code&gt; / &lt;code&gt;stale&lt;/code&gt; / &lt;code&gt;none&lt;/code&gt;), &lt;code&gt;risk.score&lt;/code&gt;, and the proxy-class signals. Verified 12 November 2026, 13:08 UTC: public DNS infrastructure here still shows a &lt;code&gt;hosting&lt;/code&gt; network attribute with score &lt;code&gt;0&lt;/code&gt;, &lt;code&gt;evidence_state: "none"&lt;/code&gt;, and &lt;code&gt;usage_type&lt;/code&gt; &lt;code&gt;IDC&lt;/code&gt; — a network fact, not a proxy verdict. The same lookup is available to agents as an MCP tool at &lt;code&gt;https://ip99.com/mcp&lt;/code&gt; (&lt;code&gt;ip_risk_lookup&lt;/code&gt;). Field reference and routing notes live on the &lt;a href="https://ip99.com/api?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;API docs&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;Try it on &lt;a href="https://ip99.com?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;ip99.com&lt;/a&gt;, keep an eye on &lt;a href="https://ip99.com/en/proxy-exits?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;this week's proxy exits&lt;/a&gt;, and wire the free OpenAPI at &lt;a href="https://ip99.com/openapi.json?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;ip99.com/openapi.json&lt;/a&gt; if you want the same fields in Postman or codegen.&lt;/p&gt;

</description>
      <category>networking</category>
      <category>security</category>
      <category>api</category>
      <category>cybersecurity</category>
    </item>
    <item>
      <title>Reading usage_type: 20 Network Scene Codes That Aren't a Risk Verdict</title>
      <dc:creator>IP99</dc:creator>
      <pubDate>Wed, 11 Nov 2026 13:21:43 +0000</pubDate>
      <link>https://dev.to/ip99/reading-usagetype-20-network-scene-codes-that-arent-a-risk-verdict-29kj</link>
      <guid>https://dev.to/ip99/reading-usagetype-20-network-scene-codes-that-arent-a-risk-verdict-29kj</guid>
      <description>&lt;p&gt;A &lt;code&gt;usage_type&lt;/code&gt; of &lt;code&gt;IDC&lt;/code&gt; is not a reason to refuse a customer, and &lt;code&gt;DYN&lt;/code&gt; is not a reason to trust one. On IP99 by ThreatHunter, &lt;code&gt;network.usage_type&lt;/code&gt; is a &lt;strong&gt;network scene code&lt;/strong&gt;: what kind of network the address sits on. It is not a risk verdict, not attacker probability, and not a synonym for &lt;code&gt;hosting&lt;/code&gt; or &lt;code&gt;proxy&lt;/code&gt;.&lt;/p&gt;

&lt;p&gt;This walkthrough stays on free, no-key &lt;code&gt;GET /v1/ip/{ip}&lt;/code&gt;. JSON samples were verified on 11 November 2026 UTC. Notes: &lt;a href="https://ip99.com?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP99&lt;/a&gt;, &lt;a href="https://ip99.com/api?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;API&lt;/a&gt;, &lt;a href="https://ip99.com/en/ip-location?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP location&lt;/a&gt;, &lt;a href="https://ip99.com/en/ip-risk-score?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP risk score&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  What &lt;code&gt;usage_type&lt;/code&gt; is (and what it is not)
&lt;/h2&gt;

&lt;p&gt;The free lookup returns a 0–100 &lt;code&gt;risk.score&lt;/code&gt; for &lt;strong&gt;freshness and strength of verifiable risk evidence&lt;/strong&gt;, not attacker probability. &lt;code&gt;risk.level&lt;/code&gt; follows the score (&lt;code&gt;0&lt;/code&gt; → &lt;code&gt;none&lt;/code&gt;, &lt;code&gt;1–29&lt;/code&gt; → &lt;code&gt;low&lt;/code&gt;, &lt;code&gt;30–69&lt;/code&gt; → &lt;code&gt;medium&lt;/code&gt;, &lt;code&gt;70–100&lt;/code&gt; → &lt;code&gt;high&lt;/code&gt;). Signals appear only when they hit. Branch on &lt;code&gt;evidence_state&lt;/code&gt; (&lt;code&gt;active&lt;/code&gt; / &lt;code&gt;stale&lt;/code&gt; / &lt;code&gt;none&lt;/code&gt;) &lt;strong&gt;before&lt;/strong&gt; the score: &lt;code&gt;0&lt;/code&gt; + &lt;code&gt;none&lt;/code&gt; is a coverage gap, not a clean bill of health.&lt;/p&gt;

&lt;p&gt;&lt;code&gt;network.usage_type&lt;/code&gt; is passed through unchanged when the scene database covers the block. Official docs list &lt;strong&gt;20 codes&lt;/strong&gt; they have a localised name for. That table is not closed: an unrecognised code is not an error — treat it as unknown rather than dropping the record. A missing &lt;code&gt;usage_type&lt;/code&gt; key means unknown, not a zero. Operator names are never returned; the machine identity is &lt;code&gt;network.asn&lt;/code&gt;.&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Group&lt;/th&gt;
&lt;th&gt;Official codes&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;Access&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;DYN&lt;/code&gt; residential broadband · &lt;code&gt;MOB&lt;/code&gt; mobile · &lt;code&gt;GTW&lt;/code&gt; enterprise gateway / NAT · &lt;code&gt;WLA&lt;/code&gt; WLAN hotspot · &lt;code&gt;DDT&lt;/code&gt; dedicated egress&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Hosting / infra&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;IDC&lt;/code&gt; datacenter · &lt;code&gt;CDN&lt;/code&gt; content delivery · &lt;code&gt;DNS&lt;/code&gt; DNS service · &lt;code&gt;ANY&lt;/code&gt; anycast · &lt;code&gt;IXP&lt;/code&gt; internet exchange · &lt;code&gt;NET&lt;/code&gt; network infrastructure&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Institution&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;EDU&lt;/code&gt; education · &lt;code&gt;GOV&lt;/code&gt; government · &lt;code&gt;ORG&lt;/code&gt; organisation&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Special / unrouted&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;SAT&lt;/code&gt; satellite · &lt;code&gt;BOGON&lt;/code&gt; reserved / bogon · &lt;code&gt;AUR&lt;/code&gt; allocated but unrouted · &lt;code&gt;RNU&lt;/code&gt; routed but unused · &lt;code&gt;UND&lt;/code&gt; unallocated · &lt;code&gt;USE&lt;/code&gt; in use&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;False equivalents:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;&lt;code&gt;usage_type&lt;/code&gt; ≠ a risk score.&lt;/strong&gt; Scene is the network; &lt;code&gt;risk.signals&lt;/code&gt; and &lt;code&gt;evidence_state&lt;/code&gt; are the observations. Official triage: &lt;strong&gt;evidence_state → score/level → signal type → usage_type&lt;/strong&gt; (then &lt;code&gt;share_tier&lt;/code&gt; on Pulse).&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;&lt;code&gt;IDC&lt;/code&gt; ≠ &lt;code&gt;DNS&lt;/code&gt; ≠ &lt;code&gt;CDN&lt;/code&gt;.&lt;/strong&gt; A famous resolver may still be labelled &lt;code&gt;IDC&lt;/code&gt;; another resolver or a CDN edge may carry a more specific code.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;&lt;code&gt;IDC&lt;/code&gt; ≠ &lt;code&gt;proxy&lt;/code&gt;.&lt;/strong&gt; Public DNS and ordinary cloud sit on &lt;code&gt;IDC&lt;/code&gt; with score &lt;code&gt;0&lt;/code&gt; every day. The same scene can also host an observed proxy exit.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;&lt;code&gt;DYN&lt;/code&gt; / &lt;code&gt;MOB&lt;/code&gt; ≠ clean.&lt;/strong&gt; Most newly seen proxy-class exits sit in residential or mobile space, so a residential allowlist has no precision against that supply.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;&lt;code&gt;hosting&lt;/code&gt; / &lt;code&gt;mobile&lt;/code&gt; are attributes derived from usage type&lt;/strong&gt;, not risk verdicts. Do not ban on those slugs alone.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;Free &lt;code&gt;/v1/ip&lt;/code&gt; does &lt;strong&gt;not&lt;/strong&gt; return &lt;code&gt;share_tier&lt;/code&gt; or &lt;code&gt;evidence[]&lt;/code&gt; (Pulse / professional only). Gauge network shape from &lt;code&gt;usage_type&lt;/code&gt;: treat &lt;code&gt;MOB&lt;/code&gt; / &lt;code&gt;GTW&lt;/code&gt; as likely shared and prefer step-up over a hard IP ban.&lt;/p&gt;

&lt;h2&gt;
  
  
  A famous resolver is &lt;code&gt;IDC&lt;/code&gt;, not &lt;code&gt;DNS&lt;/code&gt;
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/8.8.8.8
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Verified 11 November 2026 UTC:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="nl"&gt;"ip"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"8.8.8.8"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"computed_at"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"2026-11-11T13:13:09Z"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"evidence_state"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"geo"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"continent"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"NA"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"country"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"US"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"region"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"California"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"city"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Mountain View"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"tz"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"America/Los_Angeles"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"network"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"asn"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;15169&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"usage_type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"IDC"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"risk"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"score"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;0&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"level"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"signals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"hosting"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Google Public DNS is a resolver, but the scene code here is &lt;code&gt;IDC&lt;/code&gt;, with &lt;code&gt;hosting&lt;/code&gt; as a &lt;strong&gt;network attribute&lt;/strong&gt;. Score &lt;code&gt;0&lt;/code&gt; and &lt;code&gt;evidence_state: none&lt;/code&gt; mean no current verifiable proxy-class evidence — not "safe," and not a reason to refuse the address.&lt;/p&gt;

&lt;p&gt;A different public resolver, &lt;code&gt;4.2.2.1&lt;/code&gt; (ASN 3356, Los Angeles), carries &lt;code&gt;DNS&lt;/code&gt; in the same snapshot: score &lt;code&gt;0&lt;/code&gt;, &lt;code&gt;evidence_state: none&lt;/code&gt;, and no &lt;code&gt;risk.signals&lt;/code&gt; key — same job, two scene codes. Also in this snapshot (all score &lt;code&gt;0&lt;/code&gt;, &lt;code&gt;none&lt;/code&gt;): &lt;code&gt;104.16.132.229&lt;/code&gt; is &lt;code&gt;CDN&lt;/code&gt; with &lt;code&gt;hosting&lt;/code&gt; (ASN 13335); &lt;code&gt;18.26.0.1&lt;/code&gt; is &lt;code&gt;EDU&lt;/code&gt; (ASN 3, Cambridge, Massachusetts). &lt;code&gt;192.0.2.1&lt;/code&gt; (TEST-NET-1) returns &lt;code&gt;usage_type: BOGON&lt;/code&gt; and no ASN; &lt;code&gt;192.168.1.1&lt;/code&gt; returns HTTP 400 &lt;code&gt;reserved_ip&lt;/code&gt; ("private or reserved address has no public risk profile").&lt;/p&gt;

&lt;h2&gt;
  
  
  Same &lt;code&gt;IDC&lt;/code&gt; scene, opposite evidence
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/178.20.30.146
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Verified 11 November 2026 UTC:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="nl"&gt;"ip"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"178.20.30.146"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"computed_at"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"2026-11-11T13:13:21Z"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"evidence_state"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"active"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"geo"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"continent"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"EU"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"country"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"FR"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"region"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Île-de-France"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"city"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Paris"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"tz"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Europe/Paris"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"network"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"asn"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;46475&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"usage_type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"IDC"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"risk"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"score"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;80&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"level"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"high"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"signals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"hosting"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"osint"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"proxy"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Same &lt;code&gt;IDC&lt;/code&gt; + &lt;code&gt;hosting&lt;/code&gt; space as the public resolver above, but the slugs here include &lt;code&gt;proxy&lt;/code&gt; and &lt;code&gt;osint&lt;/code&gt;. Branch on &lt;strong&gt;each&lt;/strong&gt; matched signal, not on the scene. Scores decay — this address scored 84 earlier the same UTC day — so re-curl for a live snapshot.&lt;/p&gt;

&lt;p&gt;Most new proxy supply is not in that IDC slice. Official API docs include a residential (&lt;code&gt;DYN&lt;/code&gt;) demo whose 17 September 2026 capture carried &lt;code&gt;proxy&lt;/code&gt; alone (last octet masked: &lt;code&gt;86.111.136.x&lt;/code&gt;, ASN 15547). An &lt;code&gt;IDC&lt;/code&gt;-only blocklist under-covers that majority.&lt;/p&gt;

&lt;h2&gt;
  
  
  What the weekly capture mix says
&lt;/h2&gt;

&lt;p&gt;Between &lt;strong&gt;4–10 November 2026 (UTC)&lt;/strong&gt;, IP99 captured &lt;strong&gt;10,561,845&lt;/strong&gt; new proxy-class exit addresses (&lt;strong&gt;+2%&lt;/strong&gt; vs the previous 7 days). By usage type: &lt;strong&gt;80.0% residential&lt;/strong&gt;, &lt;strong&gt;12.1% mobile&lt;/strong&gt;, &lt;strong&gt;3.9% gateway/NAT&lt;/strong&gt;, &lt;strong&gt;3.2% datacenter&lt;/strong&gt;, &lt;strong&gt;0.5% education&lt;/strong&gt;. &lt;strong&gt;59&lt;/strong&gt; upstream sources. Figures generated 11 November 2026, 12:36 UTC. Source: &lt;a href="https://ip99.com/en/proxy-exits?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;proxy-exits&lt;/a&gt;. Aggregates only; no address lists.&lt;/p&gt;

&lt;p&gt;That is first-party proxy-class observation (&lt;code&gt;proxy&lt;/code&gt;, VPN, dial-up pool, cloud phone, or hijacked shared proxy), newly seen in the window, not a global census. More than &lt;strong&gt;92%&lt;/strong&gt; sit in residential or mobile space. &lt;code&gt;usage_type&lt;/code&gt; tells you where the address lives, and how wide a ban would reach, not whether the session is abusive.&lt;/p&gt;

&lt;h2&gt;
  
  
  A practical branch (evidence_state first)
&lt;/h2&gt;

&lt;p&gt;Generic guidance (not a contract): score &lt;code&gt;0&lt;/code&gt; → do not stamp "clean"; &lt;code&gt;1–29&lt;/code&gt; → observe; &lt;code&gt;30–69&lt;/code&gt; → captcha / throttle / second factor; &lt;code&gt;70–100&lt;/code&gt; → block only when business cost allows, and not at building scale. Score meaning: &lt;a href="https://ip99.com/en/ip-risk-score?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP risk score&lt;/a&gt;.&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;Read &lt;code&gt;evidence_state&lt;/code&gt;, then &lt;code&gt;risk.signals&lt;/code&gt;, then &lt;code&gt;network.usage_type&lt;/code&gt; and &lt;code&gt;network.asn&lt;/code&gt;.&lt;/li&gt;
&lt;li&gt;On &lt;code&gt;MOB&lt;/code&gt; or &lt;code&gt;GTW&lt;/code&gt;, assume a wide NAT. Prefer CAPTCHA, rate limits, or a lower weight — one IP ban can be one building or a carrier pool.&lt;/li&gt;
&lt;li&gt;On &lt;code&gt;DYN&lt;/code&gt;, do not allowlist the scene. Check whether &lt;code&gt;proxy&lt;/code&gt;, &lt;code&gt;dialup_pool&lt;/code&gt;, &lt;code&gt;vpn&lt;/code&gt;, or &lt;code&gt;hijacked_proxy&lt;/code&gt; is &lt;code&gt;active&lt;/code&gt;.&lt;/li&gt;
&lt;li&gt;On &lt;code&gt;IDC&lt;/code&gt;, &lt;code&gt;CDN&lt;/code&gt;, or &lt;code&gt;DNS&lt;/code&gt; with only &lt;code&gt;hosting&lt;/code&gt; (or no signals) and &lt;code&gt;evidence_state: none&lt;/code&gt; or &lt;code&gt;stale&lt;/code&gt;, do not treat the scene as guilt. &lt;code&gt;IDC&lt;/code&gt; plus &lt;code&gt;active&lt;/code&gt; &lt;code&gt;proxy&lt;/code&gt; / &lt;code&gt;vpn&lt;/code&gt; is a different playbook.&lt;/li&gt;
&lt;li&gt;If &lt;code&gt;hijacked_proxy&lt;/code&gt; is present: challenge; do not ban the account (the owner is the victim).&lt;/li&gt;
&lt;li&gt;Never mark score &lt;code&gt;0&lt;/code&gt; + &lt;code&gt;evidence_state: none&lt;/code&gt; as "trusted clean." Re-query at decision time and store &lt;code&gt;computed_at&lt;/code&gt;. Treat an unrecognised or missing &lt;code&gt;usage_type&lt;/code&gt; as unknown.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;Anonymous quota is &lt;strong&gt;500 lookups/day&lt;/strong&gt; per client IP (60/min); a free logged-in key raises the daily cap to &lt;strong&gt;5,000/day&lt;/strong&gt;. Check remaining with &lt;code&gt;GET https://ip99.com/v1/me&lt;/code&gt; or the &lt;code&gt;X-Daily-*&lt;/code&gt; headers. OpenAPI: &lt;a href="https://ip99.com/openapi.json?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com/openapi.json&lt;/a&gt;. Pricing: &lt;a href="https://ip99.com/en/pricing?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com/en/pricing&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;Do &lt;strong&gt;not&lt;/strong&gt; treat IP99 scores as identical to other ThreatHunter datasets — this article only claims what free &lt;code&gt;/v1/ip&lt;/code&gt; and the cited product pages document.&lt;/p&gt;

&lt;h2&gt;
  
  
  Try it
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip
curl https://ip99.com/v1/ip/8.8.8.8
curl https://ip99.com/v1/ip/4.2.2.1
curl https://ip99.com/v1/ip/178.20.30.146
curl https://ip99.com/v1/me
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;IP99 is a ThreatHunter product. Use the free lookup to read &lt;strong&gt;network scene&lt;/strong&gt; (&lt;code&gt;usage_type&lt;/code&gt;) separately from &lt;strong&gt;current risk evidence&lt;/strong&gt; — before an &lt;code&gt;IDC&lt;/code&gt; blocklist drops a public resolver or a residential allowlist waves through proxy supply. Start at &lt;a href="https://ip99.com?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com&lt;/a&gt;; read &lt;a href="https://ip99.com/en/ip-location?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP location&lt;/a&gt;, &lt;a href="https://ip99.com/en/proxy-exits?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;proxy-exits&lt;/a&gt;, and &lt;a href="https://ip99.com/api?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com/api&lt;/a&gt;.&lt;/p&gt;

</description>
      <category>security</category>
      <category>networking</category>
      <category>cybersecurity</category>
      <category>api</category>
    </item>
    <item>
      <title>Is This IP a Proxy? Reading the proxy Signal Without Blanket Blocks</title>
      <dc:creator>IP99</dc:creator>
      <pubDate>Wed, 11 Nov 2026 10:36:25 +0000</pubDate>
      <link>https://dev.to/ip99/is-this-ip-a-proxy-reading-the-proxy-signal-without-blanket-blocks-1jfj</link>
      <guid>https://dev.to/ip99/is-this-ip-a-proxy-reading-the-proxy-signal-without-blanket-blocks-1jfj</guid>
      <description>&lt;p&gt;Blanket-blocking every address tagged &lt;code&gt;proxy&lt;/code&gt; treats an exit observation as a person-level verdict. On IP99 by ThreatHunter, &lt;code&gt;proxy&lt;/code&gt; means something narrower: &lt;strong&gt;this address is currently being used as a proxy exit&lt;/strong&gt; — often because it appeared in proxy supply being sold. That is not a static commercial VPN exit, a dial-up rotation pool, or a malware-hijacked household, and not a reason to hard-ban a whole building that shares the egress.&lt;/p&gt;

&lt;p&gt;This walkthrough stays on free, no-key &lt;code&gt;GET /v1/ip/{ip}&lt;/code&gt;. JSON samples were verified on 11 November 2026 UTC; run the curls yourself for a live snapshot. Notes: &lt;a href="https://ip99.com?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP99&lt;/a&gt;, &lt;a href="https://ip99.com/api?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;API&lt;/a&gt;, &lt;a href="https://ip99.com/en/detect-proxy-vpn?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;detect proxy / VPN&lt;/a&gt;, &lt;a href="https://ip99.com/en/ip-risk-score?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP risk score&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  What &lt;code&gt;proxy&lt;/code&gt; is (and what it is not)
&lt;/h2&gt;

&lt;p&gt;IP99's free lookup returns a 0–100 &lt;code&gt;risk.score&lt;/code&gt; for &lt;strong&gt;freshness and strength of verifiable risk evidence&lt;/strong&gt;, not attacker probability. &lt;code&gt;risk.level&lt;/code&gt; follows the score (&lt;code&gt;0&lt;/code&gt; → &lt;code&gt;none&lt;/code&gt;, &lt;code&gt;1–29&lt;/code&gt; → &lt;code&gt;low&lt;/code&gt;, &lt;code&gt;30–69&lt;/code&gt; → &lt;code&gt;medium&lt;/code&gt;, &lt;code&gt;70–100&lt;/code&gt; → &lt;code&gt;high&lt;/code&gt;). Signals appear only when they hit. Always branch on &lt;code&gt;evidence_state&lt;/code&gt; (&lt;code&gt;active&lt;/code&gt; / &lt;code&gt;stale&lt;/code&gt; / &lt;code&gt;none&lt;/code&gt;) &lt;strong&gt;before&lt;/strong&gt; the score: score &lt;code&gt;0&lt;/code&gt; + &lt;code&gt;none&lt;/code&gt; is a coverage gap, not a clean bill of health. Official API wording: &lt;code&gt;proxy&lt;/code&gt; = &lt;strong&gt;this address is currently being used as a proxy exit&lt;/strong&gt; (often from observed proxy supply).&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;slug&lt;/th&gt;
&lt;th&gt;Meaning&lt;/th&gt;
&lt;th&gt;Typical disposition&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;proxy&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Address currently used as a proxy exit&lt;/td&gt;
&lt;td&gt;Step-up or block by business cost; do not ban a whole building&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;vpn&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Static commercial VPN exit&lt;/td&gt;
&lt;td&gt;Product decision; often step-up, not refuse&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;dialup_pool&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Dial-up / redial rotation pool&lt;/td&gt;
&lt;td&gt;Supply-side rotation (often DYN/MOB)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;hijacked_proxy&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Malware-hijacked egress; owner is the victim&lt;/td&gt;
&lt;td&gt;Challenge, do not ban the account&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;osint&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Listed by a public intelligence source&lt;/td&gt;
&lt;td&gt;Step-up; not a synonym for &lt;code&gt;proxy&lt;/code&gt;
&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;
&lt;code&gt;hosting&lt;/code&gt; / &lt;code&gt;mobile&lt;/code&gt;
&lt;/td&gt;
&lt;td&gt;Network attributes&lt;/td&gt;
&lt;td&gt;
&lt;strong&gt;Not&lt;/strong&gt; risk verdicts alone&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;crawler&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Search crawler egress&lt;/td&gt;
&lt;td&gt;Benign; does not score&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;False equivalents:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;&lt;code&gt;proxy&lt;/code&gt; ≠ "this user is an attacker."&lt;/strong&gt; The score is evidence freshness and strength, not person-level attribution. Shared egress means one ban can be one building.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;&lt;code&gt;proxy&lt;/code&gt; ≠ &lt;code&gt;vpn&lt;/code&gt;.&lt;/strong&gt; &lt;code&gt;vpn&lt;/code&gt; is a static commercial VPN exit, often in IDC / &lt;code&gt;hosting&lt;/code&gt; space and stable for weeks. &lt;code&gt;proxy&lt;/code&gt; is observed exit use, including rotating residential supply.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;&lt;code&gt;proxy&lt;/code&gt; ≠ &lt;code&gt;dialup_pool&lt;/code&gt;.&lt;/strong&gt; A dial-up pool is the same address being redialed for new IPs — its own slug, often on DYN or MOB.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;&lt;code&gt;proxy&lt;/code&gt; ≠ &lt;code&gt;hijacked_proxy&lt;/code&gt;.&lt;/strong&gt; That slug means compromise; the account holder is typically the victim, so challenge rather than auto-ban.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;&lt;code&gt;hosting&lt;/code&gt; on IDC ≠ &lt;code&gt;proxy&lt;/code&gt;.&lt;/strong&gt; Public DNS and ordinary cloud servers sit in hosting space every day with score &lt;code&gt;0&lt;/code&gt; and no proxy-class evidence.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;Free &lt;code&gt;/v1/ip&lt;/code&gt; does &lt;strong&gt;not&lt;/strong&gt; return &lt;code&gt;share_tier&lt;/code&gt;, &lt;code&gt;evidence[]&lt;/code&gt;, historical &lt;code&gt;?at=&lt;/code&gt;, &lt;code&gt;honeypot_scan&lt;/code&gt;, &lt;code&gt;honeypot_credential&lt;/code&gt;, or &lt;code&gt;ipv6_range&lt;/code&gt; (Pulse / professional only). On the free tier, lean on &lt;code&gt;evidence_state&lt;/code&gt;, score/level, signal type, &lt;code&gt;usage_type&lt;/code&gt;, and your own velocity / device checks.&lt;/p&gt;

&lt;h2&gt;
  
  
  Hosting without a proxy vs a datacenter proxy exit
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/8.8.8.8
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Verified 11 November 2026 UTC:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="nl"&gt;"ip"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"8.8.8.8"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"computed_at"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"2026-11-11T09:58:38Z"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"evidence_state"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"geo"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
    &lt;span class="nl"&gt;"continent"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"NA"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"country"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"US"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"region"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"California"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"city"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Mountain View"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"tz"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"America/Los_Angeles"&lt;/span&gt;
  &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"network"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"asn"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;15169&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"usage_type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"IDC"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"risk"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
    &lt;span class="nl"&gt;"score"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;0&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"level"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"signals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"hosting"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;
  &lt;span class="p"&gt;}&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;&lt;code&gt;hosting&lt;/code&gt; here is a &lt;strong&gt;network attribute&lt;/strong&gt; on public DNS. &lt;code&gt;evidence_state: none&lt;/code&gt; with score &lt;code&gt;0&lt;/code&gt; means no verifiable proxy-class evidence on file — not "safe," and &lt;strong&gt;not&lt;/strong&gt; &lt;code&gt;proxy&lt;/code&gt;. "Block all IDC" would refuse this address for the wrong reason.&lt;/p&gt;

&lt;p&gt;Now a Paris datacenter address whose live lookup carries &lt;code&gt;proxy&lt;/code&gt; (alongside &lt;code&gt;hosting&lt;/code&gt; and &lt;code&gt;osint&lt;/code&gt;):&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/178.20.30.146
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Verified 11 November 2026 UTC:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="nl"&gt;"ip"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"178.20.30.146"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"computed_at"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"2026-11-11T09:58:39Z"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"evidence_state"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"active"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"geo"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
    &lt;span class="nl"&gt;"continent"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"EU"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"country"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"FR"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"region"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Île-de-France"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"city"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Paris"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"tz"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Europe/Paris"&lt;/span&gt;
  &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"network"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nl"&gt;"asn"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;46475&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nl"&gt;"usage_type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"IDC"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="nl"&gt;"risk"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
    &lt;span class="nl"&gt;"score"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;84&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"level"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"high"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="nl"&gt;"signals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"hosting"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"osint"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"proxy"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;
  &lt;span class="p"&gt;}&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Same IDC + &lt;code&gt;hosting&lt;/code&gt; space as many VPN exits, but the slugs here are &lt;code&gt;proxy&lt;/code&gt; and &lt;code&gt;osint&lt;/code&gt;, not &lt;code&gt;vpn&lt;/code&gt;. Branch on &lt;strong&gt;each&lt;/strong&gt; matched signal. Scores decay — a documented October capture of this address scored 99 — so re-curl for a live snapshot.&lt;/p&gt;

&lt;p&gt;Most new proxy supply is not in that IDC slice. The official API docs include a residential (&lt;code&gt;DYN&lt;/code&gt;) demo whose documented capture carried &lt;code&gt;proxy&lt;/code&gt; alone (last octet masked: &lt;code&gt;86.111.136.x&lt;/code&gt;, ASN 15547, CH / Valais / Sitten; score &lt;code&gt;86&lt;/code&gt; / &lt;code&gt;high&lt;/code&gt; in the 17 September 2026 sample). A hosting-only blocklist under-covers that majority.&lt;/p&gt;

&lt;h2&gt;
  
  
  What the weekly capture mix says
&lt;/h2&gt;

&lt;p&gt;Between &lt;strong&gt;4–10 November 2026 (UTC)&lt;/strong&gt;, IP99 captured &lt;strong&gt;10,561,845&lt;/strong&gt; new proxy-class exit addresses (&lt;strong&gt;+2%&lt;/strong&gt; vs the previous 7 days). By usage type: &lt;strong&gt;80.0% residential&lt;/strong&gt; (8,445,664), &lt;strong&gt;12.1% mobile&lt;/strong&gt; (1,280,337), &lt;strong&gt;3.9% gateway/NAT&lt;/strong&gt; (412,561), &lt;strong&gt;3.2% datacenter&lt;/strong&gt; (338,371). &lt;strong&gt;59&lt;/strong&gt; upstream sources. Figures generated 11 November 2026, 09:36 UTC. Source: &lt;a href="https://ip99.com/en/proxy-exits?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;proxy-exits&lt;/a&gt;. Aggregates only; no address lists.&lt;/p&gt;

&lt;p&gt;That count is first-party proxy-class observation (&lt;code&gt;proxy&lt;/code&gt;, VPN, dial-up pool, cloud phone, or hijacked shared proxy) — newly seen in the window, not a global census, same IP counted once. More than &lt;strong&gt;92%&lt;/strong&gt; of newly seen exits sit in residential or mobile space, so a datacenter blocklist or a blanket &lt;code&gt;proxy&lt;/code&gt; ban over-punishes shared egress and still under-covers most new supply.&lt;/p&gt;

&lt;h2&gt;
  
  
  A practical branch (evidence_state first)
&lt;/h2&gt;

&lt;p&gt;Official triage order: &lt;strong&gt;evidence_state → score/level → signal type → usage_type&lt;/strong&gt;. Generic guidance (not a contract): score &lt;code&gt;0&lt;/code&gt; → do not stamp "clean"; &lt;code&gt;1–29&lt;/code&gt; → observe; &lt;code&gt;30–69&lt;/code&gt; → captcha / throttle / second factor; &lt;code&gt;70–100&lt;/code&gt; → block only when business cost allows, and not at building scale on shared egress. Score meaning: &lt;a href="https://ip99.com/en/ip-risk-score?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP risk score&lt;/a&gt;.&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;Read &lt;code&gt;evidence_state&lt;/code&gt;, then &lt;code&gt;risk.signals&lt;/code&gt; and &lt;code&gt;network.usage_type&lt;/code&gt;.&lt;/li&gt;
&lt;li&gt;If &lt;code&gt;proxy&lt;/code&gt; is &lt;code&gt;active&lt;/code&gt;: treat it as a current exit, not attacker probability. Prefer step-up plus velocity / device checks over hard-banning shared building egress. Branch separately if &lt;code&gt;vpn&lt;/code&gt;, &lt;code&gt;dialup_pool&lt;/code&gt;, or &lt;code&gt;hijacked_proxy&lt;/code&gt; also hit.&lt;/li&gt;
&lt;li&gt;If &lt;code&gt;proxy&lt;/code&gt; is &lt;code&gt;stale&lt;/code&gt;: weak, expired feature only (score &lt;code&gt;0&lt;/code&gt;). The address may already be back with an ordinary subscriber — do not apply an old exit as a fresh ban.&lt;/li&gt;
&lt;li&gt;
&lt;code&gt;hosting&lt;/code&gt; on IDC without &lt;code&gt;proxy&lt;/code&gt; / &lt;code&gt;vpn&lt;/code&gt; / &lt;code&gt;dialup_pool&lt;/code&gt; is a network attribute, not a proxy verdict. &lt;code&gt;vpn&lt;/code&gt; without &lt;code&gt;proxy&lt;/code&gt; is a static commercial VPN exit, a different playbook than rotating residential supply.&lt;/li&gt;
&lt;li&gt;If &lt;code&gt;hijacked_proxy&lt;/code&gt; is also present: challenge; do not ban the account (the owner is the victim).&lt;/li&gt;
&lt;li&gt;Never mark score &lt;code&gt;0&lt;/code&gt; + &lt;code&gt;evidence_state: none&lt;/code&gt; as "trusted clean." Re-query at decision time and store &lt;code&gt;computed_at&lt;/code&gt; rather than caching a verdict for days.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;Anonymous quota is &lt;strong&gt;500 lookups/day&lt;/strong&gt; per client IP (60/min); a free logged-in key raises the daily cap to &lt;strong&gt;5,000/day&lt;/strong&gt;. Check remaining with &lt;code&gt;GET https://ip99.com/v1/me&lt;/code&gt; or the &lt;code&gt;X-Daily-*&lt;/code&gt; headers. OpenAPI: &lt;a href="https://ip99.com/openapi.json?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com/openapi.json&lt;/a&gt;. Pricing: &lt;a href="https://ip99.com/en/pricing?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com/en/pricing&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;Do &lt;strong&gt;not&lt;/strong&gt; treat IP99 scores as identical to other ThreatHunter datasets — this article only claims what free &lt;code&gt;/v1/ip&lt;/code&gt; and the cited product pages document.&lt;/p&gt;

&lt;h2&gt;
  
  
  Try it
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip
curl https://ip99.com/v1/ip/8.8.8.8
curl https://ip99.com/v1/ip/178.20.30.146
curl https://ip99.com/v1/me
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;IP99 is a ThreatHunter product. Use the free lookup to separate a &lt;strong&gt;current proxy exit&lt;/strong&gt; (&lt;code&gt;proxy&lt;/code&gt;) from commercial VPN static exits, dial-up rotation pools, hijacked-victim egress, ordinary hosting, and leftover labels after an address changed hands — before a blanket block bans a customer who shares the building's egress. Start at &lt;a href="https://ip99.com?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com&lt;/a&gt;; read &lt;a href="https://ip99.com/en/detect-proxy-vpn?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;detect proxy / VPN&lt;/a&gt;, &lt;a href="https://ip99.com/en/proxy-exits?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;proxy-exits&lt;/a&gt;, and &lt;a href="https://ip99.com/api?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com/api&lt;/a&gt;.&lt;/p&gt;

</description>
      <category>security</category>
      <category>api</category>
      <category>proxy</category>
      <category>cybersecurity</category>
    </item>
    <item>
      <title>This Week's Proxy Exits: 10.6M New Addresses, Still Almost All Off Datacenter Space</title>
      <dc:creator>IP99</dc:creator>
      <pubDate>Wed, 11 Nov 2026 10:06:59 +0000</pubDate>
      <link>https://dev.to/ip99/this-weeks-proxy-exits-106m-new-addresses-still-almost-all-off-datacenter-space-5827</link>
      <guid>https://dev.to/ip99/this-weeks-proxy-exits-106m-new-addresses-still-almost-all-off-datacenter-space-5827</guid>
      <description>&lt;p&gt;Datacenter blocklists still look like the default proxy defence in a lot of fraud stacks. The addresses showing up in proxy supply this week were still mostly not in datacenters. The datacenter slice eased from the 3.6% in the last posted recap, though it remains above the flat ~2% band of mid-to-late October.&lt;/p&gt;

&lt;p&gt;Between &lt;strong&gt;4–10 November 2026 (UTC)&lt;/strong&gt;, the &lt;a href="https://ip99.com?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP99&lt;/a&gt; risk feed captured &lt;strong&gt;10,561,845&lt;/strong&gt; proxy exit addresses it had not seen in the previous 7 days (&lt;strong&gt;+2%&lt;/strong&gt; vs the previous 7 days). Source page: &lt;a href="https://ip99.com/en/proxy-exits?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;ip99.com/en/proxy-exits&lt;/a&gt; (generated 11 November 2026, 09:36 UTC). The public page rolls daily over the last 7 complete UTC days; this is the window it reports today, continuing the series after the 29 October–4 November recap (the two windows overlap on 4 November).&lt;/p&gt;

&lt;p&gt;Against that last posted week (&lt;strong&gt;10,576,402&lt;/strong&gt; addresses), volume is essentially unchanged (&lt;strong&gt;−0.1%&lt;/strong&gt;). The page's official &lt;strong&gt;+2%&lt;/strong&gt; compares 4–10 November with 28 October–3 November, not with the overlapping posted window.&lt;/p&gt;

&lt;h2&gt;
  
  
  What those addresses sit on
&lt;/h2&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Usage type&lt;/th&gt;
&lt;th&gt;Addresses&lt;/th&gt;
&lt;th&gt;Share&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;Residential broadband&lt;/td&gt;
&lt;td&gt;8,445,664&lt;/td&gt;
&lt;td&gt;80.0%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Mobile network&lt;/td&gt;
&lt;td&gt;1,280,337&lt;/td&gt;
&lt;td&gt;12.1%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Gateway / NAT egress&lt;/td&gt;
&lt;td&gt;412,561&lt;/td&gt;
&lt;td&gt;3.9%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Datacenter&lt;/td&gt;
&lt;td&gt;338,371&lt;/td&gt;
&lt;td&gt;3.2%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Education network&lt;/td&gt;
&lt;td&gt;51,606&lt;/td&gt;
&lt;td&gt;0.5%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Unclassified&lt;/td&gt;
&lt;td&gt;33,306&lt;/td&gt;
&lt;td&gt;0.3%&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;About &lt;strong&gt;63%&lt;/strong&gt; were rated high-risk at the moment of capture. The rest still carried a proxy-class signal at a lower risk level.&lt;/p&gt;

&lt;p&gt;IP99 counts an address as "captured" when it enters the feed with a proxy-class signal (&lt;code&gt;proxy&lt;/code&gt;, &lt;code&gt;VPN&lt;/code&gt;, dial-up pool, cloud phone, or hijacked shared proxy) and was &lt;strong&gt;not&lt;/strong&gt; in the live 7-day risk mirror. Addresses that go quiet and return can be counted again. The page describes what &lt;strong&gt;59&lt;/strong&gt; upstream collection sources delivered — not the total number of proxies in the world. An address is counted once per window. No address lists are published.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why 3.2% still is not "mostly datacenter"
&lt;/h2&gt;

&lt;p&gt;If your defence is mainly an ASN or hosting blocklist, you are still mostly looking past the problem. Residential broadband alone is &lt;strong&gt;80.0%&lt;/strong&gt; of this week's capture window. Mobile adds another &lt;strong&gt;12.1%&lt;/strong&gt;. Together that is over &lt;strong&gt;92%&lt;/strong&gt; of newly seen proxy exits sitting in access networks where real customers also live.&lt;/p&gt;

&lt;p&gt;The datacenter row moved from &lt;strong&gt;3.6%&lt;/strong&gt; in the last posted recap (29 Oct–4 Nov) to &lt;strong&gt;3.2%&lt;/strong&gt; here, and from about &lt;strong&gt;2.0%&lt;/strong&gt; in the 26 Oct–1 Nov window a fortnight back. That is a real mix shift worth watching, but it does not flip the headline. A hosting-only blocklist still misses the majority of the supply-side traffic that shows up as residential or mobile exits.&lt;/p&gt;

&lt;p&gt;Usage type answers "what kind of network is this," not "is this address currently sold as an exit." Rotating residential proxies and dial-up pools are designed to sit in DYN / MOB space. Detecting them needs supply-side observation with an age attached, not a permanent network-type label. Product notes: &lt;a href="https://ip99.com/en/rotating-residential-proxies?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;rotating residential proxies&lt;/a&gt;, &lt;a href="https://ip99.com/en/detect-proxy-vpn?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;detect proxy / VPN&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  Which networks show up most
&lt;/h2&gt;

&lt;p&gt;Top ASNs in this week's capture window (by distinct address count):&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;ASN&lt;/th&gt;
&lt;th&gt;Addresses&lt;/th&gt;
&lt;th&gt;Share&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;AS4134&lt;/td&gt;
&lt;td&gt;949,981&lt;/td&gt;
&lt;td&gt;9.0%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;AS4837&lt;/td&gt;
&lt;td&gt;728,248&lt;/td&gt;
&lt;td&gt;6.9%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;AS9808&lt;/td&gt;
&lt;td&gt;355,509&lt;/td&gt;
&lt;td&gt;3.4%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;AS6057&lt;/td&gt;
&lt;td&gt;268,061&lt;/td&gt;
&lt;td&gt;2.5%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;AS36903&lt;/td&gt;
&lt;td&gt;207,084&lt;/td&gt;
&lt;td&gt;2.0%&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;AS4134, AS4837, and AS9808 remain the three largest contributors, as they have in recent recaps. AS36903 enters the top five this window; AS55836, which held fifth place in the 29 Oct–4 Nov recap (2.2%), sits lower in this week's table (1.6%, seventh). Large access networks appearing here are a clue about where proxy supply is being sold from — not a claim that every customer on those networks is a proxy. Ban a whole ASN and you take out legitimate home and mobile users. Prefer per-address evidence at decision time. ASN pages for deeper network mix: &lt;a href="https://ip99.com/en/asn/?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;ip99.com/en/asn/&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  Day-to-day shape
&lt;/h2&gt;

&lt;p&gt;Captured addresses (new to the prior 7-day mirror that day), with how many were high-risk at capture:&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Day (UTC)&lt;/th&gt;
&lt;th&gt;Captured&lt;/th&gt;
&lt;th&gt;High-risk at capture&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-04&lt;/td&gt;
&lt;td&gt;1,752,678&lt;/td&gt;
&lt;td&gt;1,169,732&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-05&lt;/td&gt;
&lt;td&gt;1,493,146&lt;/td&gt;
&lt;td&gt;946,402&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-06&lt;/td&gt;
&lt;td&gt;1,544,553&lt;/td&gt;
&lt;td&gt;943,591&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-07&lt;/td&gt;
&lt;td&gt;1,517,140&lt;/td&gt;
&lt;td&gt;919,244&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-08&lt;/td&gt;
&lt;td&gt;1,415,823&lt;/td&gt;
&lt;td&gt;897,828&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-09&lt;/td&gt;
&lt;td&gt;1,371,503&lt;/td&gt;
&lt;td&gt;883,005&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;2026-11-10&lt;/td&gt;
&lt;td&gt;1,467,002&lt;/td&gt;
&lt;td&gt;940,244&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;Volume rose slightly week-over-week (&lt;strong&gt;+2%&lt;/strong&gt; after the immediately preceding ~10.3M days), while upstream sources sat at &lt;strong&gt;59&lt;/strong&gt; (57 in the last posted recap). The firmer headline is about what those collectors delivered into the 7-day mirror, not a claim that proxy supply worldwide grew. Daily capture stayed in a band roughly &lt;strong&gt;1.37–1.75M&lt;/strong&gt; addresses, with &lt;strong&gt;4 Nov&lt;/strong&gt; the heavier day — still not a shape that makes a static list look "caught up."&lt;/p&gt;

&lt;h2&gt;
  
  
  How to use this without over-blocking
&lt;/h2&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Branch on evidence, not network type.&lt;/strong&gt; Residential is where your customers live; do not hard-block on DYN / MOB alone.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Step up, don't stop.&lt;/strong&gt; A fresh proxy-class signal on signup, a coupon claim, or a login from a new device justifies a second factor or a delay — not an immediate ban of a whole building's shared egress.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Read the age.&lt;/strong&gt; &lt;code&gt;evidence_state: "stale"&lt;/code&gt; means the address was in supply before and is not now; treat it as a weak prior. Re-query at decision time rather than caching a verdict for days. Score &lt;code&gt;0&lt;/code&gt; with &lt;code&gt;evidence_state: "none"&lt;/code&gt; is not a clean bill of health either — it means no current verifiable evidence is on file.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Keep the timestamp.&lt;/strong&gt; Store &lt;code&gt;computed_at&lt;/code&gt; with the decision so a disputed block can be traced back to when the observation was made.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;IP99 by ThreatHunter publishes the free weekly aggregate so you can see the mix without buying a feed. The score on a single lookup is the freshness and strength of verifiable risk evidence — not a probability that the address "is an attacker." Do not treat IP99 scores as identical to every other ThreatHunter dataset you may have seen elsewhere; this article only claims what the public proxy-exits page and free &lt;code&gt;/v1/ip&lt;/code&gt; document.&lt;/p&gt;

&lt;h2&gt;
  
  
  Check one yourself
&lt;/h2&gt;

&lt;p&gt;Free, no signup:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/8.8.8.8
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Look at &lt;code&gt;evidence_state&lt;/code&gt; (&lt;code&gt;active&lt;/code&gt; / &lt;code&gt;stale&lt;/code&gt; / &lt;code&gt;none&lt;/code&gt;), &lt;code&gt;risk.score&lt;/code&gt;, and the proxy-class signals. Public DNS infrastructure often shows a &lt;code&gt;hosting&lt;/code&gt; network attribute with score &lt;code&gt;0&lt;/code&gt; and &lt;code&gt;evidence_state: "none"&lt;/code&gt; — that is a network fact, not a proxy verdict. The same lookup is available to agents as an MCP tool at &lt;code&gt;https://ip99.com/mcp&lt;/code&gt; (&lt;code&gt;ip_risk_lookup&lt;/code&gt;). Field reference and routing notes live on the &lt;a href="https://ip99.com/api?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;API docs&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;Try it on &lt;a href="https://ip99.com?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;ip99.com&lt;/a&gt;, keep an eye on &lt;a href="https://ip99.com/en/proxy-exits?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;this week's proxy exits&lt;/a&gt;, and wire the free OpenAPI at &lt;a href="https://ip99.com/openapi.json?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;ip99.com/openapi.json&lt;/a&gt; if you want the same fields in Postman or codegen.&lt;/p&gt;

</description>
      <category>security</category>
      <category>cybersecurity</category>
      <category>networking</category>
      <category>api</category>
    </item>
    <item>
      <title>Don't Ban on a Public List Alone: Reading osint on IP99</title>
      <dc:creator>IP99</dc:creator>
      <pubDate>Sun, 08 Nov 2026 13:13:45 +0000</pubDate>
      <link>https://dev.to/ip99/dont-ban-on-a-public-list-alone-reading-osint-on-ip99-1n16</link>
      <guid>https://dev.to/ip99/dont-ban-on-a-public-list-alone-reading-osint-on-ip99-1n16</guid>
      <description>&lt;p&gt;Public IP blocklists feel definitive: if an address showed up in open-source intelligence, refuse it. That treats a listing as a permanent identity. Addresses change hands; community lists refresh on a slower clock. On IP99 by ThreatHunter, &lt;code&gt;osint&lt;/code&gt; names something narrower: the address was &lt;strong&gt;listed by a public intelligence source&lt;/strong&gt;. It is a risk tag, not a probability that the person behind the IP is an attacker, and not a reason to skip the rest of the triage.&lt;/p&gt;

&lt;p&gt;This walkthrough stays on free, no-key &lt;code&gt;GET /v1/ip/{ip}&lt;/code&gt;. JSON samples below are documented product shapes from IP99 docs and prior verified captures — run the curls yourself for a live snapshot. We use a documented capture that already carries &lt;code&gt;osint&lt;/code&gt; alongside other slugs; we do &lt;strong&gt;not&lt;/strong&gt; invent a sample IP that carries &lt;code&gt;osint&lt;/code&gt; alone. Notes: &lt;a href="https://ip99.com?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP99&lt;/a&gt;, &lt;a href="https://ip99.com/api?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;API&lt;/a&gt;, &lt;a href="https://ip99.com/en/detect-proxy-vpn?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;detect proxy / VPN&lt;/a&gt;, &lt;a href="https://ip99.com/en/ip-risk-score?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP risk score&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  What &lt;code&gt;osint&lt;/code&gt; is (and what it is not)
&lt;/h2&gt;

&lt;p&gt;IP99's free lookup returns a 0–100 &lt;code&gt;risk.score&lt;/code&gt; for &lt;strong&gt;freshness and strength of verifiable risk evidence&lt;/strong&gt;, not attacker probability. &lt;code&gt;risk.level&lt;/code&gt; follows the score (&lt;code&gt;0&lt;/code&gt; → &lt;code&gt;none&lt;/code&gt;, &lt;code&gt;1–29&lt;/code&gt; → &lt;code&gt;low&lt;/code&gt;, &lt;code&gt;30–69&lt;/code&gt; → &lt;code&gt;medium&lt;/code&gt;, &lt;code&gt;70–100&lt;/code&gt; → &lt;code&gt;high&lt;/code&gt;). Signals appear only when they hit. Always branch on &lt;code&gt;evidence_state&lt;/code&gt; (&lt;code&gt;active&lt;/code&gt; / &lt;code&gt;stale&lt;/code&gt; / &lt;code&gt;none&lt;/code&gt;) &lt;strong&gt;before&lt;/strong&gt; the score: score &lt;code&gt;0&lt;/code&gt; + &lt;code&gt;none&lt;/code&gt; is a coverage gap, not a clean bill of health.&lt;/p&gt;

&lt;p&gt;Tags that matter here:&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;slug&lt;/th&gt;
&lt;th&gt;Meaning&lt;/th&gt;
&lt;th&gt;Typical disposition&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;osint&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Listed by a public intelligence source (risk tag)&lt;/td&gt;
&lt;td&gt;Step-up; do not hard-ban on a listing alone&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;proxy&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Address currently used as a proxy exit&lt;/td&gt;
&lt;td&gt;Step-up or block by business cost&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;vpn&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Static commercial VPN exit&lt;/td&gt;
&lt;td&gt;Product decision; often step-up, not refuse&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;dialup_pool&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Dial-up / redial rotation pool&lt;/td&gt;
&lt;td&gt;Supply-side rotation (often DYN/MOB)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;hijacked_proxy&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Malware-hijacked egress; owner is the victim&lt;/td&gt;
&lt;td&gt;Challenge, do not ban the account&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;
&lt;code&gt;hosting&lt;/code&gt; / &lt;code&gt;mobile&lt;/code&gt;
&lt;/td&gt;
&lt;td&gt;Network attributes&lt;/td&gt;
&lt;td&gt;
&lt;strong&gt;Not&lt;/strong&gt; risk verdicts alone&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;crawler&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Search crawler egress&lt;/td&gt;
&lt;td&gt;Benign; does not score&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;Official API wording: &lt;code&gt;osint&lt;/code&gt; = &lt;strong&gt;listed by a public intelligence source&lt;/strong&gt; (open-source intelligence). It can appear in &lt;code&gt;risk.signals&lt;/code&gt; and, on Pulse, as an &lt;code&gt;evidence[].risk_type&lt;/code&gt; row. The docs do not name upstream list vendors, and this article does not invent them.&lt;/p&gt;

&lt;p&gt;False equivalents:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;&lt;code&gt;osint&lt;/code&gt; ≠ "this user is an attacker."&lt;/strong&gt; The score measures evidence freshness and strength. A listing is an observation about an address, not attribution of a person. Shared exits put many unrelated users behind one IP.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;&lt;code&gt;osint&lt;/code&gt; ≠ &lt;code&gt;proxy&lt;/code&gt; / &lt;code&gt;vpn&lt;/code&gt; / &lt;code&gt;dialup_pool&lt;/code&gt;.&lt;/strong&gt; Those tags come from observed proxy-class use or commercial VPN infrastructure. &lt;code&gt;osint&lt;/code&gt; means a public source listed the address. They can appear together; they are not synonyms.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;&lt;code&gt;hosting&lt;/code&gt; on IDC ≠ &lt;code&gt;osint&lt;/code&gt;.&lt;/strong&gt; Public DNS and ordinary cloud servers sit in hosting space every day with score &lt;code&gt;0&lt;/code&gt; and no public-intel hit.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;&lt;code&gt;crawler&lt;/code&gt; is benign&lt;/strong&gt; and does not score — do not fold a search crawler into an OSINT block.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;Free &lt;code&gt;/v1/ip&lt;/code&gt; does &lt;strong&gt;not&lt;/strong&gt; return &lt;code&gt;share_tier&lt;/code&gt; (Pulse only). On free tier, lean on &lt;code&gt;evidence_state&lt;/code&gt;, score/level, signal type, &lt;code&gt;usage_type&lt;/code&gt;, plus your own velocity / device checks.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why public listings go stale
&lt;/h2&gt;

&lt;p&gt;IP99's detect-proxy/VPN guide puts it plainly: two clocks run at different speeds. &lt;strong&gt;The address changes hands in minutes; the list that describes it is refreshed in days.&lt;/strong&gt; The gap is where false positives live — a real customer gets the address a list still describes as yesterday's exit.&lt;/p&gt;

&lt;p&gt;That is why IP99 treats every answer as a judgement that expires, not a permanent label. When &lt;code&gt;osint&lt;/code&gt; is present, read &lt;code&gt;evidence_state&lt;/code&gt; and &lt;code&gt;computed_at&lt;/code&gt; with it. &lt;code&gt;active&lt;/code&gt; means fresh scoring evidence; &lt;code&gt;stale&lt;/code&gt; means it was seen but has aged out (score &lt;code&gt;0&lt;/code&gt;, weak feature only); &lt;code&gt;none&lt;/code&gt; means never seen — unknown, not safe.&lt;/p&gt;

&lt;p&gt;Official corollary from the same page: no source is complete. Proxy supply is a market; every vendor sees a slice of it. Treat any single feed, IP99 included, as one input among several. A public listing is useful context, not a substitute for first-party &lt;code&gt;proxy&lt;/code&gt; / &lt;code&gt;dialup_pool&lt;/code&gt; evidence.&lt;/p&gt;

&lt;h2&gt;
  
  
  Documented contrast: hosting without a listing vs listing plus proxy vs proxy without osint
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/8.8.8.8?utm_source&lt;span class="o"&gt;=&lt;/span&gt;devto&amp;amp;utm_medium&lt;span class="o"&gt;=&lt;/span&gt;article&amp;amp;utm_campaign&lt;span class="o"&gt;=&lt;/span&gt;ip99
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Documented product sample (IP99 docs / prior verified capture):&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="nl"&gt;"ip"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"8.8.8.8"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="nl"&gt;"computed_at"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"2026-10-09T10:22:26Z"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="nl"&gt;"evidence_state"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="nl"&gt;"geo"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="nl"&gt;"continent"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"NA"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="nl"&gt;"country"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"US"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="nl"&gt;"region"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"California"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="nl"&gt;"city"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"Mountain View"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="nl"&gt;"tz"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"America/Los_Angeles"&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="p"&gt;},&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="nl"&gt;"network"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="nl"&gt;"asn"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="mi"&gt;15169&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="nl"&gt;"usage_type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"IDC"&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;},&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="nl"&gt;"risk"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="nl"&gt;"score"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="mi"&gt;0&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="nl"&gt;"level"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"none"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="nl"&gt;"signals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"hosting"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="w"&gt;
&lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="w"&gt;
&lt;/span&gt;&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;&lt;code&gt;hosting&lt;/code&gt; here is a &lt;strong&gt;network attribute&lt;/strong&gt; on public DNS. &lt;code&gt;evidence_state: none&lt;/code&gt; with score &lt;code&gt;0&lt;/code&gt; means no verifiable proxy-class or public-intel evidence on file — not "safe," and &lt;strong&gt;not&lt;/strong&gt; &lt;code&gt;osint&lt;/code&gt;. Blocking every address that "looks listed" or "looks like a datacenter" would refuse this one for the wrong reason.&lt;/p&gt;

&lt;p&gt;Now a Paris datacenter address whose documented capture included &lt;code&gt;osint&lt;/code&gt; together with &lt;code&gt;proxy&lt;/code&gt; and &lt;code&gt;hosting&lt;/code&gt;:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/178.20.30.146?utm_source&lt;span class="o"&gt;=&lt;/span&gt;devto&amp;amp;utm_medium&lt;span class="o"&gt;=&lt;/span&gt;article&amp;amp;utm_campaign&lt;span class="o"&gt;=&lt;/span&gt;ip99
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Documented product sample (IP99 docs / prior verified capture):&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="nl"&gt;"ip"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"178.20.30.146"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="nl"&gt;"computed_at"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"2026-10-09T10:23:02Z"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="nl"&gt;"evidence_state"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"active"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="nl"&gt;"geo"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="nl"&gt;"continent"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"EU"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="nl"&gt;"country"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"FR"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="nl"&gt;"region"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"Île-de-France"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="nl"&gt;"city"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"Paris"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="nl"&gt;"tz"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"Europe/Paris"&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="p"&gt;},&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="nl"&gt;"network"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="nl"&gt;"asn"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="mi"&gt;46475&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="nl"&gt;"usage_type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"IDC"&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;},&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="nl"&gt;"risk"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="nl"&gt;"score"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="mi"&gt;99&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="nl"&gt;"level"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"high"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="nl"&gt;"signals"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"hosting"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"osint"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"proxy"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="w"&gt;
&lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="w"&gt;
&lt;/span&gt;&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;This is the documented shape we can show: &lt;code&gt;osint&lt;/code&gt; matched, and so did &lt;code&gt;proxy&lt;/code&gt;. Branch on &lt;strong&gt;each&lt;/strong&gt; matched slug. &lt;code&gt;proxy&lt;/code&gt; means observed proxy egress; &lt;code&gt;osint&lt;/code&gt; means a public intelligence source listed it; &lt;code&gt;hosting&lt;/code&gt; is only the network attribute. Do not collapse the three into "bad IP, ban." We are not inventing a solo-&lt;code&gt;osint&lt;/code&gt; address — if a live lookup later drops a slug, that is expected, because scores decay.&lt;/p&gt;

&lt;p&gt;Most new proxy supply is not in that IDC slice. API docs demo for a residential proxy egress &lt;strong&gt;without&lt;/strong&gt; &lt;code&gt;osint&lt;/code&gt;:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip/86.111.136.23?utm_source&lt;span class="o"&gt;=&lt;/span&gt;devto&amp;amp;utm_medium&lt;span class="o"&gt;=&lt;/span&gt;article&amp;amp;utm_campaign&lt;span class="o"&gt;=&lt;/span&gt;ip99
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Documented product sample (API docs demo): score &lt;code&gt;86&lt;/code&gt;, level &lt;code&gt;high&lt;/code&gt;, signals &lt;code&gt;["proxy"]&lt;/code&gt;, &lt;code&gt;usage_type&lt;/code&gt; &lt;code&gt;DYN&lt;/code&gt;, ASN &lt;code&gt;15547&lt;/code&gt;, geo CH / Valais / Sitten. Re-curl for a live snapshot — scores decay with evidence age. First-party proxy observation can exist without a public listing, just as a listing can linger after the address changed hands.&lt;/p&gt;

&lt;h2&gt;
  
  
  What the weekly capture mix says
&lt;/h2&gt;

&lt;p&gt;Between &lt;strong&gt;1–7 November 2026 (UTC)&lt;/strong&gt;, IP99 captured &lt;strong&gt;10,728,498&lt;/strong&gt; new proxy-class exit addresses (+3% WoW). By usage type: &lt;strong&gt;79.2% residential&lt;/strong&gt; (8,498,288), &lt;strong&gt;12.3% mobile&lt;/strong&gt; (1,316,500), &lt;strong&gt;4.0% gateway/NAT&lt;/strong&gt; (425,657), &lt;strong&gt;3.7% datacenter&lt;/strong&gt; (391,719). &lt;strong&gt;58&lt;/strong&gt; upstream sources. Figures generated 8 November 2026 UTC. Source: &lt;a href="https://ip99.com/en/proxy-exits?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;proxy-exits&lt;/a&gt;. Aggregates only; no address lists.&lt;/p&gt;

&lt;p&gt;That weekly count is first-party proxy-class observation (proxy, VPN, dial-up pool, cloud phone, or hijacked shared proxy) — &lt;strong&gt;not&lt;/strong&gt; a count of &lt;code&gt;osint&lt;/code&gt; listings. The mix still matters: most new exits sit in residential space. Trusting only public lists, or only blocking IDC, under-covers that majority and over-punishes leftover listings after an address changed hands.&lt;/p&gt;

&lt;h2&gt;
  
  
  A practical branch (evidence_state first)
&lt;/h2&gt;

&lt;p&gt;Official triage order: &lt;strong&gt;evidence_state → score/level → signal type → usage_type&lt;/strong&gt;. Generic guidance (not a contract): score &lt;code&gt;0&lt;/code&gt; → do not stamp "clean"; &lt;code&gt;1–29&lt;/code&gt; → observe; &lt;code&gt;30–69&lt;/code&gt; → captcha / throttle / second factor; &lt;code&gt;70–100&lt;/code&gt; → block only when business cost allows. Score meaning: &lt;a href="https://ip99.com/en/ip-risk-score?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;IP risk score&lt;/a&gt;.&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;1. Read evidence_state (active / stale / none).
2. Read risk.signals and network.usage_type.
3. If osint is present and evidence_state is active:
     - Treat as "listed by a public intelligence source."
     - Do NOT treat the listing as attacker probability or as a person-level verdict.
     - Prefer step-up + velocity / device / session checks over a hard ban on the listing alone.
     - If proxy / vpn / dialup_pool are also present, branch on those slugs too — they are different observations.
4. If osint is present and evidence_state is stale:
     - Weak, expired feature only (score 0). Do not apply an old list hit as a fresh ban.
5. Else if hosting on IDC WITHOUT osint / proxy / vpn:
     - Network attribute only — do not invent a public-intel verdict.
6. Else if proxy on DYN or MOB without osint:
     - First-party proxy supply, not a public listing — different from osint-only reasoning.
7. If hijacked_proxy is also present: challenge; do not ban the account (owner is the victim).
8. Never mark score 0 + evidence_state none as "trusted clean."
9. Re-query at decision time; do not cache an osint verdict for days.
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Anonymous quota is &lt;strong&gt;500 lookups/day&lt;/strong&gt; per client IP (60/min); a free logged-in key raises the daily cap to &lt;strong&gt;5,000/day&lt;/strong&gt;. Check remaining with &lt;code&gt;GET https://ip99.com/v1/me&lt;/code&gt;?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99 or the &lt;code&gt;X-Daily-*&lt;/code&gt; headers. OpenAPI: &lt;a href="https://ip99.com/openapi.json?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com/openapi.json?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99&lt;/a&gt;. Pricing: &lt;a href="https://ip99.com/en/pricing?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com/en/pricing?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;Do &lt;strong&gt;not&lt;/strong&gt; treat IP99 scores as identical to other ThreatHunter datasets — this article only claims what free &lt;code&gt;/v1/ip&lt;/code&gt; and the cited English product pages / OpenAPI document.&lt;/p&gt;

&lt;h2&gt;
  
  
  Try it
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://ip99.com/v1/ip?utm_source&lt;span class="o"&gt;=&lt;/span&gt;devto&amp;amp;utm_medium&lt;span class="o"&gt;=&lt;/span&gt;article&amp;amp;utm_campaign&lt;span class="o"&gt;=&lt;/span&gt;ip99
curl https://ip99.com/v1/ip/8.8.8.8?utm_source&lt;span class="o"&gt;=&lt;/span&gt;devto&amp;amp;utm_medium&lt;span class="o"&gt;=&lt;/span&gt;article&amp;amp;utm_campaign&lt;span class="o"&gt;=&lt;/span&gt;ip99
curl https://ip99.com/v1/ip/178.20.30.146?utm_source&lt;span class="o"&gt;=&lt;/span&gt;devto&amp;amp;utm_medium&lt;span class="o"&gt;=&lt;/span&gt;article&amp;amp;utm_campaign&lt;span class="o"&gt;=&lt;/span&gt;ip99
curl https://ip99.com/v1/ip/86.111.136.23?utm_source&lt;span class="o"&gt;=&lt;/span&gt;devto&amp;amp;utm_medium&lt;span class="o"&gt;=&lt;/span&gt;article&amp;amp;utm_campaign&lt;span class="o"&gt;=&lt;/span&gt;ip99
curl https://ip99.com/v1/me?utm_source&lt;span class="o"&gt;=&lt;/span&gt;devto&amp;amp;utm_medium&lt;span class="o"&gt;=&lt;/span&gt;article&amp;amp;utm_campaign&lt;span class="o"&gt;=&lt;/span&gt;ip99
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;IP99 is a ThreatHunter product. Use the free lookup to separate a &lt;strong&gt;public intelligence listing&lt;/strong&gt; (&lt;code&gt;osint&lt;/code&gt;) from first-party proxy observation, commercial VPN exits, ordinary hosting, and leftover labels after an address changed hands — before a stale list hit bans a customer who just received the IP. Start at &lt;a href="https://ip99.com?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99&lt;/a&gt;; read &lt;a href="https://ip99.com/en/detect-proxy-vpn?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;detect proxy / VPN&lt;/a&gt;, &lt;a href="https://ip99.com/en/proxy-exits?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;proxy-exits&lt;/a&gt;, and &lt;a href="https://ip99.com/api?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99" rel="noopener noreferrer"&gt;https://ip99.com/api?utm_source=devto&amp;amp;utm_medium=article&amp;amp;utm_campaign=ip99&lt;/a&gt;.&lt;/p&gt;

</description>
      <category>cybersecurity</category>
      <category>security</category>
      <category>api</category>
      <category>osint</category>
    </item>
  </channel>
</rss>
