Analytics
See every lookup your API keys made, who called, and which IPs looked risky.
The Analytics page lists the lookups made with your API keys and shows which of the IPs you asked about were VPNs, proxies, Tor exits, hosting addresses or on a threat list.
What is stored
For each looked-up IP we keep one record with:
- the time, the endpoint (
/v1/lookup/:ipor/v1/lookup/bulk), the response status and how long we took - which API key was used, or Dashboard if the lookup came from the IP lookup page
- the IP address the request came from (the caller) and the IP you asked about
- the classification we returned for it: flags, fraud score and recommended action
- the first 120 characters of the user agent
A bulk call is stored as one record per IP, all sharing the same request ID.
We do not store request bodies, API keys, JWTs, cookies or query strings.
How long it is kept
| Plan | Request history |
|---|---|
| Free | 7 days |
| Starter | 30 days |
| Business | 1 year |
| Enterprise | 1 year by default |
Records are deleted automatically when the window ends. If you move to a plan with a shorter window, older records are deleted and the new window applies right away. If you move to a longer one, the records you already have are kept for the longer window. Records that were already deleted can't be brought back.
What you see
Date range
The button at the top right sets the range. It opens quick ranges (Today, Yesterday, Last 7 days, Last 30 days, Last 90 days, This month, Last month) next to a two-month calendar. To pick your own dates, click the first day, then the last day. The page opens on Today.
Days are UTC days, and the last day is included. You can pick any dates, but you only see what your plan keeps (7 days on Free, 30 on Starter, 1 year on Business and Custom). If your selection reaches past that, the page shows the part your plan keeps and says so.
Cards
Seven cards show counts for the range you picked: Total lookups, VPN, Proxy (residential proxy included), Tor, Hosting, Threat flags and Privacy relay. Each lookup counts once per card, so a lookup that matches two threat lists is one Threat flags lookup.
Each card shows how the count changed against the previous period of equal length: the same number of days straight before your range. For Last 7 days that is the 7 days before. For Today it is yesterday up to the same time of day, so a half-finished day isn't compared with a whole one. A green badge means a move in the good direction and a red one the other way, so more VPN lookups show red. New means the count was zero in the previous period, so there is no percentage to show. If the previous period starts before the history your plan keeps, the cards show no comparison.
For ranges longer than 31 days the cards come from daily totals: Threat flags is then an upper bound, because one lookup with several threat flags is counted once per flag.
Filters
Filters narrows the cards and the request log together. Active filters appear as chips you can remove with the x.
- Source: All, API key or Dashboard.
- Flag: VPN, Proxy, Residential proxy, Tor, Relay, Hosting, Threat, Crawler or Bogon. Threat covers abusive, botnet, spammer, scanner and spoofed crawler. Crawler covers verified and spoofed crawlers.
- Status: Success, Not found, Rate limited or Server error.
- IP: a full IPv4 or IPv6 address, matched exactly against the IP you looked up.
Filters need a range of 31 days or less for the cards. On a longer range the log is still filtered, and the cards show all lookups.
Request log and IP tables
The Request log, newest first, shows when each lookup happened (how long ago, plus the exact time), its source, the IP you looked up, the caller IP, country, ASN and organisation, flags, fraud score and action, status and latency. The eye button opens the full record.
Top flagged IPs and Who called your API sit below the log. An unfamiliar caller address can mean a leaked key, so rotate that key from API keys. They use at most the last 31 days of your range.
Flagged does not mean fraud. Hosting IPs include your own servers and legitimate bots, and privacy relays are usually real people.
What the numbers mean
- Total lookups: lookups made with your API keys (and from the dashboard IP lookup page) in the range you picked. A bulk call counts once per IP, the same as your quota.
- Success rate: the share of lookups that returned data. 404 means the IP isn't in our database yet; 429 means you hit your daily limit; 5xx means something failed on our side.
- Flagged: lookups where the IP you asked about is a VPN, proxy, residential proxy, Tor exit, privacy relay, hosting or datacenter address, or is on a threat list (abusive, botnet, spammer, scanner, spoofed crawler, bogon). Flagged doesn't mean fraud: hosting IPs include your own servers and legitimate bots.
- VPN: the IP belongs to a commercial VPN exit.
- Proxy and residential proxy: traffic was relayed through this IP. Residential proxies run on home or mobile lines, which makes them harder to spot by network alone.
- Tor: the IP is a Tor exit node.
- Privacy relay: a privacy relay such as iCloud Private Relay. Usually a real person who wants privacy, not an attacker.
- Hosting: the IP is in a datacenter or cloud network, not a home or mobile connection.
- Threat flags: the IP appears on abuse, botnet, spam or scanner lists, or claims to be a search crawler without matching that crawler's published ranges.
- Fraud score and action: our 0 to 100 risk score for the IP and the action it suggests: allow, review, step up (ask for extra verification) or block.
- Unique IPs: distinct IP addresses you looked up in the range.
- Who called your API: the addresses your requests came from, usually your own servers. An unfamiliar one can mean a leaked key.
- Latency: time from when your request reached us to when we sent the response, measured on our side. p95 means 95% of requests were this fast or faster.
Who can see it
Analytics are shown in the dashboard only. API keys can't read them, so a leaked key can't be used to read your traffic.