navigator
webgl
canvas
timezone
Identity Architecture

Identity Is
Entropy
Coherence

A browser fingerprint is the combined observable profile created by device characteristics, engine behavior, rendering traits, API surfaces, and how all correlate over time. Anti-bot systems detect inconsistency, not incorrect values.

identity_coherence_check , runtime
Platform consistency
PASS
Renderer coherence
PASS
API surface alignment
PASS
Temporal stability
PASS
Distribution fit
PASS
Coherence Score 0.997
String-Only Identity
Incoherent
Common Failures

"Plausible" Is
Multidimensional

1

"If values look right, it's fine"

Detection systems measure more than exposed strings. Many high-signal traits are derived from real engine behavior, correlated across multiple APIs, and sensitive to version distribution.

If your fingerprint claims "Windows + Chrome X" but behaves differently under deeper measurement, you are not "close." You are distinct.
2

"Freeze and reuse forever"

Fingerprints age. The web platform shifts constantly: browser versions roll forward, graphics stacks change, API surfaces expand, distributions change by region and device class.

A "perfect" frozen profile becomes an outlier over time, its own risk signal.
Observable Surfaces 200+ parameters
Device & OS 32 params
platform cpuClass memory cores
Engine Behavior 48 params
V8 timing math precision date parsing regex engine
Rendering & Graphics 56 params
canvas entropy webgl vendor fonts color depth
API Surfaces 72 params
permissions plugins mimeTypes featurePolicy
Cross-Surface Correlation

Anti-bot systems evaluate consistency across all surfaces simultaneously, not individual values in isolation.

At-Scale Ingestion

Millions of Real-World
Fingerprints Per Month

Realism is distributional. Undetect ingests millions of fingerprints monthly across Windows, macOS, and Android, capturing depth across versions, builds, device classes, regions, and hardware variability.

Windows
4.2M
Monthly fingerprints
10 11 Server 2022
macOS
2.8M
Monthly fingerprints
13.x 14.x Apple Silicon
Android
3.5M
Monthly fingerprints
12 13 14
Distribution Depth Real-world variance captured
847
OS Versions
2,341
Browser Builds
156
Device Classes
94
Regions
412
GPU Variants
fingerprint_resolver , constraint engine
// Minimal constraints from user
const constraints = {
platform: 'windows',
region: 'us-east',
screenClass: 'laptop',
deviceClass: 'corporate'
};
// Resolve to full coherent profile
const profile = await resolver.resolve(constraints);
[RESOLVE] Matching against 10.7M profiles...
[RESOLVE] Platform constraint: 4.2M candidates
[RESOLVE] Region constraint: 892K candidates
[RESOLVE] Screen coherence: 234K candidates
[RESOLVE] Distribution fit: 12K optimal
[RESOLVE] Selected: Windows 11 / Chrome 124 / Intel i5
// Full profile ready before launch
await
browser.launch({ profile });
$
Resolution Time
<50ms
Pre-Launch Resolution

From Minimal Hints
To Coherent Identity

Most customers do not want to specify 200+ parameters. They want to say "Windows, US, typical laptop" and receive a cohesive full profile prior to launch.

Late-Stage Patching Risk

Patching identity after browser start risks early measurements capturing wrong state, contradictions between surfaces, and subtle execution mismatches.

Pre-Launch Coherence

Identity is coherent from session start. No early leakage, no surface contradictions, no execution pathway mismatches.

Operational Identity

Persistent.
Evolving.
Continuous.

Some workflows require stable identity across time. Others need freshness. Undetect supports both with an operational identity model built around coherence and continuity.

Persistent Identities

For returning-user semantics, account-based sessions, and long-lived agent personas. Persisted fingerprints tied to specific identities enable controlled reuse over time, not just "reuse the same JSON" but an operational identity model with session continuity.

Account Warmth Returning User Long-Lived Personas

Session Sync

A fingerprint alone is not identity continuity. Session Sync pulls and backs up cookies, local/session storage, permissions, and session artifacts, giving you deterministic continuity instead of fragile custom state handling.

Full cookie jar
Storage state
Permissions
Fingerprint binding

Automatic Evolution

Fingerprints become stale as platforms shift. Undetect updates fingerprints over time each time a browser launches, maintaining coherence, evolving plausibly, avoiding frozen outliers, preserving continuity without violating realism.

Goal: Stay "in distribution" as the platform evolves
Optional SDK

Mirror Your Real
User Population

Some teams must match their own end users precisely, when the "correct" profile is not generic realism but exact alignment to a known device base.

Deployable SDK

Capture device profiles from your own application users under your control.

Precise Mirroring

Align agent identities to actual customer device distributions.

Highest Fidelity

The most accurate approach when exact population match is required.

SDK Integration Flow
Your Application
SDK embedded in client app
Fingerprint Ingestion
Anonymous profile capture
Distribution Analysis
Your user base profiled
Agent Alignment
Your agents match your users
Foundation, Not Premium

Unlimited Fingerprints.
No Extra Cost.

Fingerprints are foundational, not a premium line item. Per-fingerprint pricing pushes customers toward unsafe shortcuts. The correct unit of value is successful, reliable operation, not a JSON object.

Fingerprints
0
Per-Identity Fees
100%
Included
Proof Under Your Constraints

Validate Coherence
In Your Environment

Fingerprints should be validated in the real threat environment: your targets, your regions, your workflow patterns. We demonstrate pre-launch resolution, continuity via persisted identities, controlled evolution, and stable success rates across repeated runs.

Pre-Launch
Coherence Resolution
Persistent
Identity Continuity
Evolving
Distribution Fit
Stable
Repeated Success