An anti api abuse project by Qyzar
Our goal: observe, learn, and evolve defenses against bots and automated attacks—so legitimate traffic stays fast and abuse gets harder over time.
Step 1
Using over 360+ sources of information we fingerprint the user's network, hardware, and device using a series of tests and calculations—creating the base of our fingerprinting journey.
Step 2
We check over 160 anomalies and tell-tale bot signs. We don't only rely on entropy—we score based on past browser data, IP geo data, ASN suspicion, and thousands of other sites' records of the user.
Step 3
We silently collect entropy, creating a larger image of what the user is doing and a summarization of their behaviour.
Step 4
When risk is high, we escalate with an in-house captcha system—no third-party widgets. Levels include depth-analysis captchas, image identification, and object-location tasks in the scene, ramping difficulty so bots face a steeper wall than real users.
Step 5
As fingerprinting is updated, each user's device information develops into more accurate, compounded data. Our merger bot walks every device, finds similarities that may not have been obvious before, and tightens the picture over time.
Tap when you're in view—we sample APIs locally. Charts and pills summarize what the raw numbers mean—same card chrome as the rest of the marketing pages.
Snapshot
—
Pointer, scroll, and key volume plus first-input delay—real sessions look uneven; bots and background tabs often skew ratios.
Logical cores, RAM hint, touch capacity, WebGL GPU strings, and a silent audio render hash—together they cluster like a device family.
Display size, viewport, DPR, a paint hash—and a live mini snapshot of this page as your browser painted it when you tapped Ready.
Timezone, languages, platform, UA, and client hints—honest browsers line up; mismatches often mean privacy tools or spoofing.
Cookies, storage, the webdriver flag, PDF viewer, and theme—cheap checks before heavier proof or friction.
Same idea as the timeline and pricing blocks: many small fields, one scored picture—here you see the raw surface.
Efficient storage keeps our unit economics low. We apply advanced compression to fingerprint payloads so stored representations stay compact without sacrificing the fields our models need. Lower storage and transfer overhead means we can offer broad coverage at price points many competing services cannot match—without cutting corners on signal quality. Beyond price, tiers differ by how many concurrent sessions you can run and how deep the signal stack goes. Choose the level that matches your traffic and compliance posture, and upgrade when you outgrow the limits.
No credit card required
Up to 10 active sessions
Validate integration paths, run proofs of concept, and ship small pilots without committing to a subscription.
Billed monthly · cancel anytime
Up to 20,000 active sessions
Scale to production traffic with a high session ceiling and more traffic rules for live apps.
Billed monthly · cancel anytime
Unlimited sessions
Our deepest signal stack for teams that rely on entropy and long-horizon behavior for risk, fraud, and trust decisions.
