v2.3-GOLD Base-Zero, with receipts

IPQS over-flags.
v2.3 returns the math.

IPQualityScore is built around aggressive heuristics that often torch residential ranges. CandycornDB v2.3 starts at 0, applies asnResidentialBonus -10, and returns a signed scoreReasons receipt with every call. Less collateral. More signal.

No Credit Card Required
Get API Key Read the Docs first

100 calls per day on the Developer tier. Free forever.

CLEAN COMCAST IP
$

curl candycorndb.com/api/public/lookup?ip=73.14.58.201

// IPQS often scores this 75+ on a residential range. We score 0.
{
"ip": "73.14.58.201",
"score": 0,
"scoreVersion": "v2.3-base-zero",
"asnType": "residential",
"isProxy": false,
"isp": "Comcast",
"scoreReasons": [
{ "component": "asnResidentialBonus", "delta": -10 }
]
}

Feature parity, side by side

An aggressive black box cannot answer the questions a v2.3 receipt does.

Capability CandycornDB v2.3 IPQualityScore
Residential False-Positive Rate Low: asnResidentialBonus -10 High: aggressive heuristics
Subnet-Aware Scoring /24 CIDR clusters, +25 Per-IP only
Wholesale Landlord ID Clouvider, M247, Tesonet Generic ASN tagging
Base-Zero Transparency scoreReasons[] on every call Black-box composite score
Pricing Public: $0 / $49 / $299 Tiered, complex, sales-led
Cached Lookup p99 < 50ms Variable

Aggressive black box vs live refinery pipeline

A composite score with no decomposition cannot defend a block. A receipt can.

Aggressive · Opaque

IPQualityScore

Heuristic stack tuned for recall. Composite scores hide the math.

Step 1
IP fed into composite model
Multiple internal heuristics applied. Weighting is proprietary and not exposed.
Step 2
Residential ranges aggressively flagged
A handful of bad reports on a Comcast /16 can color clean neighbors. No asnResidentialBonus exists.
Step 3
You get a fraud score, no math
A 75 returns. You cannot explain to support why a real customer was blocked.
Step 4
Tune your threshold by trial and error
Block at 75? 80? 90? Each move trades false positives for false negatives blindly.
Live · Auditable

CandycornDB v2.3 Refinery

Pipeline model with explicit deltas, residential bonuses, and signed receipts.

Step 1
Base-Zero start
Score begins at 0. The default assumption is "innocent until proven otherwise."
Step 2
Trust bonuses for clean signals
Residential -10, mobile -5, gated by !isProxy. Real customers stay green.
Step 3
Penalties only when justified
Hosting +15, proxy +20, Tor +45, cluster +25, weighted abuse +25. Each delta has a name.
Step 4
Audit receipt returned
scoreReasons[] ships every component. Set thresholds with confidence: score >= 40 || isProxy.

The audit receipt IPQS cannot return

A clean Comcast IP that IPQS would often score 75+ on aggressive heuristics. CandycornDB scores 0 with a signed receipt.

// GET /api/public/lookup?ip=73.14.58.201 // Real residential customer. IPQS may flag aggressively. v2.3 returns the math. { "ip": "73.14.58.201", "score": 0, "scoreVersion": "v2.3-base-zero", "country": "US", "asn": "AS7922", "asnType": "residential", "isProxy": false, "isp": "Comcast", "scoreReasons": [ { "component": "asnResidentialBonus", "delta": -10, "detail": "comcast matched RESIDENTIAL_KEYWORDS, !isProxy" } ] } // Math: 0 (base) + (-10) trust bonus = -10, clamped to 0. // Risk band: Low. Action: pass through unfrictioned.

A score of 0 you can defend. A score of 75 with no breakdown is a guess that costs you customers.

No Credit Card Required

Stop trusting stale databases.
Start auditing your traffic.

100 calls per day on the Developer tier. v2.3 Base-Zero engine. scoreReasons on every call.

Get API Key Read the Docs first