98.6% OF PUMP.FUN TOKENS ARE RUGS

See the rug
before it pulls.

DrainBrain’s 5‑model AI ensemble predicts rug pulls with F1 = 1.000 accuracy. Wallet DNA fingerprinting. Plain‑English verdicts. Not another score — actual protection.

0+

RISK FEATURES

0

AI MODELS

0

RUG STAGES

0

MCP SERVERS

175,000+ Tokens AnalyzedF1 = 1.000 Accuracy5 AI ModelsSub-Second ResponseJupiter + Helius + Phantom

HOW IT WORKS

Three steps to rug immunity

STEP 01

Paste Token Address

Drop any Solana token mint. Free users get 10 scans per day, no signup required.

Works in browser, Telegram, or REST API

STEP 02

DrainBrain Analyzes

5-model ML ensemble scores 80+ features. Wallet DNA mapped. Honeypot simulated. Rug stage classified.

Proprietary 5-model ensemble

STEP 03

Get Your Verdict

Plain-English AI verdict. What-If loss simulation. One-click swap out via Jupiter if it's risky.

Not a score — an actual explanation

DRAINBRAIN™ ENGINE

The AI that reads the blockchain’s mind

Not just another token checker. DrainBrain is a proprietary multi-model ensemble trained on 175,000+ labeled tokens — the largest Solana security dataset ever assembled.

drainbrain --scan
>

5-MODEL ENSEMBLE

Proprietary ML Architecture

Five specialized ML models including NVIDIA NIM plus battle-tested heuristic rules. Weighted ensemble catches patterns no single model can.

F1 = 1.000validation accuracy

WALLET DNA

Graph Intelligence

Community detection, reputation propagation, neural graph embeddings, and trust scoring. Every wallet gets a forensic fingerprint.

Multi-dimneural embeddings

TEMPORAL TRACKING

5-Stage Rug Sequence

Temporal ML model watches behavior across multiple time windows. Detects the progression from setup to pull in real time.

4time windows

HONEYPOT SIM

Jupiter Buy+Sell Test

Simulates a buy and sell through Jupiter to verify the token is actually sellable. Catches honeypots before you spend a cent.

< 1ssimulation time

WHAT-IF ENGINE

Loss Simulation

"If the deployer removes liquidity, your $100 becomes $0.03." See exact dollar impact before the rug, not after.

$0.03simulated outcome

AI VERDICTS

Plain-English Explanations

No raw numbers. Grok AI generates a verdict explaining exactly why a token is risky, referencing wallet history and on-chain evidence.

1clear answer

MCP SERVER

Agent-Native API

Connect any AI agent via Model Context Protocol. Pay per call with x402 USDC micropayments. No subscription, no API key needed.

$0.01per scan

CHROME EXTENSION

Pre-Swap Interception

Automatic risk verdict before every trade on Jupiter, Raydium, Pump.fun, and more. Catches rugs before you sign the transaction.

11DEX sites

PUBLISHED ACCURACY

We don’t say “trust us.” We publish the numbers.

No other Solana security tool publishes precision, recall, and F1 scores. Our models are validated against 175,000+ labeled tokens from multiple curated datasets.

1.000

F1 Score

Harmonic mean of precision & recall

100%

Precision

Zero false positives on validation set

100%

Recall

Zero false negatives — every rug caught

PRICING

Protection for every level

Start free. Pay with card or USDC. No lock-in.

Free

$0
  • +10 scans per day
  • +Basic risk score
  • +Authority checks
  • +Honeypot detection
Get Started
MOST POPULAR

Pro

$29/mo
  • +100 scans per day
  • +DrainBrain AI verdicts
  • +What-If simulation
  • +Rug stage tracking
  • +Telegram scan + alerts
Start Pro

All plans accept Card or USDC on Solana

FOR DEVELOPERS

Add rug detection to your bot

One POST request. Sub-second response. The same 5-model ensemble that powers RugSlayer, available as a REST API.

drainbrain-api.sh
curl -X POST /api/drainbrain/v1/scan \
  -H "Authorization: Bearer db_live_..." \
  -H "Content-Type: application/json" \
  -d '{"mint": "TOKEN_ADDRESS"}'

# Response:
{
  "score": 87,
  "riskLevel": "CRITICAL",
  "isRug": true,
  "rugStage": 4,
  "honeypot": { "isHoneypot": false },
  "confidence": 0.97
}

Sub-second response

ML inference in under 1ms. Full pipeline under 3 seconds including on-chain data.

🔒

Free tier included

100 scans/day with a free API key. No credit card required. Upgrade for 10K+/min.

🧠

5-model ensemble

Same 5-model proprietary ensemble. Honeypot simulation, behavioral staging, risk flags.

📊

Tier-gated detail

Free: score + risk level. Pro: full breakdown, honeypot, temporal analysis, rug stage.

View API Docs & Get Free Key

WHY RUGSLAYER

Static snapshots are dead

FeatureRugSlayerRugCheck / GoPlus / SolSniffer
Behavioral prediction (pre-rug)
Published accuracy metrics
Wallet DNA fingerprinting
What-If loss simulation
5-model ML ensemble
Plain-English AI verdicts
Honeypot simulation
One-click swap out
MCP server for AI agents
Chrome extension (11 DEXs)
Telegram scan + alerts
REST API for developers
Free tier available

STOP GETTING RUGGED

98.6% of tokens are rugs.
Be in the 1.4%.

Scan your first token in under 10 seconds. Free, no signup required. DrainBrain does the rest.