Hextner

IP reputation and bot detection APIs for every request 

Hextner makes two security APIs. VerifIP is an IP reputation and fraud-scoring API that scores any IP address, email, phone number or URL from your server, and DetectBT is invisible browser bot detection whose signed verdict your server checks. Know who is really behind every single request.

Drops into the stack you already run

CloudflareVercelNext.jsNode.jsExpressFastifyNGINXDjangoLaravelGoRailsSupabasePythonDocker

What do VerifIP andDetectBT each check?

VerifIP scores the network a request came from. DetectBT checks the browser it claims to be. Each returns its own verdict with the reasons behind it, and your code decides what to do with both.

Globe, network and location icons feed the Hextner hub, which outputs a shield and a check mark.

What happens whenthe API cannot answer?

Hextner is built to sit in the request path of a production system. When VerifIP cannot answer, it says so with a 503 and no verdict. By contrast, when DetectBT cannot score a visitor, your server lets the request through, flagged. Either way, your code decides.

Runs on Cloudflare’s network, so each call is accepted at the Cloudflare location nearest your server. VerifIP’s threat data keeps its primary copy in Singapore.

The Hextner hub linked to six nodes: shield, globe, chart, clock, server rack and fingerprint.

Where do IP reputation andbot detection help?

The same two signals answer a surprising number of questions, for instance:

A sign-up form passes through Hextner: four sign-ups are accepted and one is rejected.

Signup fraud

Catch throwaway accounts at registration — disposable email domains, proxy and datacenter origins, and scripted sign-ups — before they ever cost you a seat.

A card payment passes through Hextner to one of three outcomes: allow, challenge or block.

Payment fraud

Score the network and the device at checkout, and hand the risk band straight to the payment rules you already run.

Login attempts pass through Hextner: one person is let in and the scripted attempts are rejected.

Credential stuffing

VerifIP flags the proxy, hosting and blocklisted IPs a stuffing run comes from; DetectBT flags the automated browsers that drive it.

A browser that claims to be Chrome is found to be headless by Hextner and rejected.

Scraping and bots

Headless browsers and spoofed user agents surface in the fingerprint even when the IP looks clean.

Six account profiles traced by Hextner to one device fingerprint, labelled as one person.

Multi-accounting

Account farms rotate through VPNs, proxies and datacenter ranges, and drive browsers with automation. VerifIP flags the first; DetectBT catches the second.

Built to sit inthe request path

<50ms
p50 latency, measured at our edge
99.99%
uptime SLA on Scale
~23 KB
browser script, gzipped
90 days
request-log retention, then deleted

Integrations

How does it fitinto my stack?

Plain HTTPS, plain JSON. Both products answer your server over HTTPS with JSON, so the verdict lands in your own code: log it, queue it, or feed it to the rules you already run. There are no push integrations or log exports yet.

Pricing

How muchdoes it cost?

Pay for what you score. A VerifIP request counts once whatever the verdict; invalid requests, 429s and capacity 503s are never billed. DetectBT bills one evaluation per page load, plus one for each additional token, and never bills a bot verdict.

Free
$0forever

Enough to wire it up and see real verdicts on your own traffic.

Start free
  • 10,000 requests / month
  • Both products included
  • Usage console with request logs
  • Email support · no response target
Starter
$99per month

For a product in market that needs detection in the request path.

Get Starter
  • 50,000 requests / month
  • VerifIP + DetectBT bundle
  • Bursts up to 30 requests / 10s
  • Batch endpoints · 1–100 items
  • Email support · first reply in 1 business day

Need custom volume? See all plans

Frequently askedquestions

The things engineers ask before wiring detection into a live product.

Integration is one HTTPS call from your server for VerifIP, and one script tag plus a server-side /verify call for DetectBT. The free tier is enough to try both on production traffic before you commit.

VerifIP is one HTTPS call, built to sit in the request path. If we cannot answer, you get an explicit 503 (X-VerifIP-Degraded) instead of a guessed verdict. The docs show a bounded retry followed by failing closed, and a 10-second client timeout keeps a slow network from hanging your request path. DetectBT does the opposite by default: if it cannot score a visitor, your server lets the request through, flagged.

Queried values (IP, email, phone, URL) are stored in your request logs for 90 days so you can review them in the console, then deleted. We do not build a cross-customer profile of anyone.

Yes. VerifIP and DetectBT are separate calls, priced separately. Use one, or call both on the same request and combine the two answers in your own code.

Neither product gives you a bare yes or no. Every VerifIP answer carries its signal_breakdown, the signals behind the score; DetectBT gives your server the verdict, a risk_score and the reason. Set your own thresholds in your own code.

VerifIP API requests beyond your plan's quota are refused with HTTP 429 until the quota resets or you upgrade. DetectBT never refuses your visitors: after a short grace allowance, evaluations pass through unscored. You are never charged more than your plan price.

Ready to see who isbehind your traffic?