Live · EVM, Solana & Bitcoin

Blockchain Forensics,
Automated.

Paste any wallet address: Anoxir collects its full transaction history across chains, scores risk with a multi-model ML ensemble, screens sanctions and scam registries, traces tainted funds hop by hop — and generates court-ready forensic reports in seconds.

20+ Chains supported
9 Intelligence sources
6 ML models in ensemble
Analysis Engine
v3.2 · LIVE
Risk Assessment
⚠ INVESTIGATE — Moderate Signals Detected
64
Active Signals
Threat Intelligence Match HIGH · 91%
Behavioral Pattern Anomaly MED · 67%
ML Ensemble Consensus ensemble · 74%
Regulatory Screening CLEAR
Sources active
Regulatory DB
Threat Intel
Community BL
Label Registry
7 / 7 online
Compliance-Ready · SOC 2 Roadmap · GDPR Aware · API-First · Cloud SaaS · Wyoming LLC

Platform Capabilities

Everything your team needs
to analyze risk — fast.

Anoxir collects the full on-chain history of any wallet, scores its risk with a machine-learning ensemble, screens sanctions and threat-intel registries, then packages everything into forensic-grade reports.

Multi-Hop Taint Tracing

Follow tainted funds hop by hop across the network — the taint engine tracks value through DEX swaps, cross-asset transformations and mixer detours, with evidence stamped on every hop.

Taint Engine DEX Splices Mixer Detection

Automated Forensic Reports

Generate professional-grade PDF reports in one click. Structured around Mandiant/CrowdStrike-style frameworks: facts → patterns → interpretation → conclusions.

PDF Export Chain of Custody JSON/CSV

Sanctions & Threat-Intel Screening

Every address is checked against OFAC SDN, curated blacklists, community scam registries and whitelists — a confirmed hard fact always overrides any model output.

OFAC SDN Blacklists Whitelists

Network Graph Intelligence

Interactive graph visualization of entity relationships and fund flows. Automatically identifies clusters, bridges and high-risk propagation paths — and recognizes known fraud topologies: pig butchering hubs, address-poisoning vanity addresses, honeypots, drainer kits, rug pulls, dust attacks and cash-out paths.

ML Ensemble Risk Scoring

Random Forest, Gradient Boosting, GNN and GCN models score every wallet in ensemble on each analysis. Every signal is weighted, explainable, and capped by hard facts.

Workflow

From query to report
in under 60 seconds.

Step 01

Submit your Query

Paste any wallet address — EVM chains, Solana or Bitcoin. Anoxir routes it through 9 intelligence sources and collects its full on-chain history in parallel.

Step 02

ML Analysis Runs

ML models vote in ensemble — behavioral, graph, temporal — producing a composite, explainable risk score for every address on the path.

Step 03

Export Actionable Report

Download a structured forensic PDF or raw JSON for integration into your existing compliance toolchain.

One platform, three levels

From a 10-second check
to a full investigation.

Anoxir adapts to who's asking. Basic users get an instant answer on a given address. Analysts go deep and build the full picture of a suspect operation — or formally clear an address of any red flag. And on top sits a supervised AI layer that turns raw results into plain language.

Level 1 · For everyone

Quick Address Check

Paste a wallet address and get an immediate verdict against the 9 intelligence sources — sanctions lists, scam reports, phishing registries, curated whitelists. In seconds you know whether an address appears anywhere in the known-threat landscape, before sending funds or interacting with a counterparty.

~10 seconds · no expertise needed
Level 2 · For analysts

Deep Forensic Analysis

The complete investigation: full transaction history, fund-flow graph, hop-by-hop taint tracing, and the ML ensemble scoring every address on the path. Use it to document a scam end to end — identity of the clusters, flow of the funds, destination venues — or to formally verify the opposite: that an address carries no red flag, with the evidence trail to prove it.

~60 seconds · report-grade output
Cross-cutting · Supervised AI

AI-Interpreted Results

A supervised AI layer reads the generated data — graph, scores, signals — and restates it as comprehensible text: what happened, why the score, what matters next. It assists investigations by pointing to the hop that needs attention, suggesting leads to clear a doubt, and writing the narrative sections of the exported report. It never invents facts: every sentence traces back to data.

explainable · data-bound · reviewer-supervised

Community evidence

Been scammed?
File a testimony that counts.

Anoxir doubles as a reporting channel. Victims file structured testimonies against a scam address (type of scam, prejudice amount, evidence screenshots/documents) — or, on the contrary, file a formal defense for a wrongly flagged address. Reports rated credible by the review workflow feed the threat-intel layer used by every future analysis.

  • Evidence pack encrypted client-side (you hold the password), sealed on IPFS — tamper-proof and timestamped
  • Jurisdiction tagging so reports surface to the right local authorities and community labels
  • Anti-abuse by design: daily quotas, file screening, accusation and defense submissions
What a testimony produces
Sealed evidence bundle preserved on IPFS (CID reference)
Structured report usable for authorities and lawyers
Community signal that strengthens the next analyses
A live deep-analysis of the accused address, if you want one attached
File a testimony on anoxir.io →
20+
Chains (EVM, Solana, BTC)
9
Intelligence Sources
6
ML Models (RF, GB, GNN, GCN)
100%
Explainable Results

How the scoring works

A transparent score,
not a black box.

Most risk tools give you a number you have to trust blindly. Anoxir Technologies shows its reasoning. Every assessment is built in three layers, and every signal that contributes to the final score is listed with its own confidence — so an analyst can defend the conclusion, not just quote it.

Layer 1

Identity & known facts

First we check hard, verifiable facts — regulatory lists, known-good registries and community intelligence. A confirmed fact is what drives a strong verdict, never a model guess alone.

Layer 2

Machine-learning diagnosis

An ensemble of ML models votes — decision forests, gradient boosting, graph neural networks, temporal models. Each contributes a weighted, explainable signal rather than a single opaque output.

Layer 3

Neighbourhood exposure

Finally we measure exposure to the surrounding network — clusters, bridges and propagation paths — to understand risk by association, not just the entity in isolation.

Evidence-based, not accusatory

A model on its own can flag for review, but it can never produce a maximum-risk verdict by itself. A severe conclusion always requires a hard, verifiable fact behind it.

Explainable by design

In-app panels describe each algorithm in plain language, and exports include the per-signal breakdown — so the analysis is understandable without a data-science background.

Product tour

See it in action.

A look at the analyst workspace, the relationship graph, and the reports you can export.

Anoxir analysis workspace overview
Primary screenshot — analyst workspace
place file at landing/screenshots/cockpit-overview.png
recommended 2400×1500 px, PNG
Network relationship graph
Network graph view
place file at landing/screenshots/network-graph.png
Exported forensic PDF report
Exported report (PDF)
place file at landing/screenshots/forensic-report.png
Network Graph
TARGET ⚠ flagged ⚠ suspect
Verified entity Flagged entity Unknown

Graph Intelligence

See the full picture,
not just the data point.

Anoxir Technologies' graph engine maps relationships and propagation paths across your entire dataset — automatically clustering entities, surfacing hidden bridges, and quantifying risk exposure by neighborhood.

  • Automatic node labeling with confidence scores
  • Interactive 3D graph with filtering and timeline
  • Community detection via Leiden clustering algorithm
  • Contamination exposure score by neighborhood

Pricing

Simple, transparent plans.

Start with a single one-time analysis, or scale with a subscription when your caseload grows. Prices are in euros and take effect at purchase on anoxir.io.

🚀 Launch pricing — these rates are promotional and will increase at public launch. Early testers run analyses for free on anoxir.io while the offer lasts.

Single

€4.99

One-time purchase.

  • 1 analysis
  • Core ML risk scoring
  • PDF & JSON exports
Analyze now

Starter

€40 / month

Billed monthly. Cancel anytime.

  • Up to 10 analyses / month
  • Core ML risk scoring
  • PDF & JSON exports
  • Email support
Get started
Most popular

Professional

€150 / month

Billed monthly or annually.

  • Up to 50 analyses / month
  • Full ML ensemble + graph engine
  • Network graph intelligence
  • API access
  • Priority support
Get started

Enterprise

€450 / month

Billed monthly or annually.

  • Up to 200 analyses / month
  • Full ML ensemble + graph engine
  • Network graph intelligence
  • API access + webhooks
  • Dedicated support
Talk to us

All prices are in euros and exclude applicable taxes. Subscriptions renew automatically and can be cancelled at any time from your account. See our Refund & Cancellation Policy.

The app is live — anoxir.io

Investigate your first address.

The full platform is live: paste a wallet address, get the fund-flow graph, the risk score and the forensic report. No waiting list, no sales call required — analyses are free for the first testers while launch pricing is in effect.

⛓️
Multi-chain
EVM · Solana · Bitcoin
🇺🇸
Anoxir Technologies LLC
Wyoming, USA
✉️
contact@anoxir.io
Response <48h