What does this coin scanner do?

It reads public data on newly listed tokens — price, volume, holder tables, order flow, contract details — then scores each token 0–100 against a fixed set of criteria, and states which metric each token failed on. It does not trade and does not give advice.

Where does the data come from?

DexScreener, GeckoTerminal, RugCheck, GoPlus, the public pump.fun API and the PumpPortal websocket. All open sources; no paid keys.

How often does the table update?

The server scans on a fixed schedule: pump.fun and multi-platform every 5 minutes, newly listed CoinMarketCap tokens every 5 minutes, the mid and high market-cap groups every 15 minutes, and the track-record table every 3 hours.

Does a high score mean it is worth buying?

No. The score only says that at scan time the token met preset thresholds on supply distribution, liquidity and contract safety. It does not forecast price and does not replace your own research. Most new tokens lose nearly all their value regardless of score.

Why is a token that is trending not in the table?

The filter drops tokens there is not enough data to measure — no on-chain pair yet, no readable holder table, or simply too new. Those sit in the Reserve list with the reason they failed, rather than disappearing.

Do I need an account?

No. This page is read-only: no run buttons, no editable thresholds, and everyone sees the same table the server produced.

Coin scanner — how to read the table and how scoring works

The scanner reads public data on newly listed tokens across many decentralised exchanges and chains, then scores each one against a fixed set of criteria. This page explains every number in the table.

Open the scanner

Four tables, four different stages

Each table targets one stage in a token's life and uses its own criteria, because what is worth looking at differs by stage.

Pump.fun: Tokens still on the bonding curve, before they reach a decentralised exchange. Read straight from the pump.fun API, with one-minute candles and holder tables flagged for snipers and bundling.
Multi-platform: Every launchpad and decentralised exchange across several chains. Slower, because it has to query more sources.
New on CMC: Tokens added to CoinMarketCap's new listings within the last 48 hours. Beyond market and holder data it also measures real order flow, the Telegram community, and how verifiable the project's identity is.
Mid 100k–1M and High 1M–10M: Tokens that have existed for days or longer, scored on durability rather than novelty: price held against the 30- or 90-day high, wallet count, number of pools, and which centralised exchanges already list them.

Reading the columns

Score: 0–100 total. 70 and above is well worth a close look, 55–70 worth a look, 40–55 ordinary, below 40 weak.
Market cap: Total token value at the current trading price.
Vol 30m: Money traded in the last 30 minutes. High and widely spread suggests real participants.
+30m: Percentage price change over 30 minutes.
Holders: Number of wallets holding the token.
Top 10: What share the ten largest wallets hold. This is the single most important column: the lower it is, the more evenly supply is spread.
Dev: What share the token's creator still holds.
Age: How many minutes or hours since the token launched.

How the score is calculated

The total is the sum of nine components, each with a fixed weight. The table below is the weighting currently in use.

The nine components of the total score
Component	Weight	What it measures
Supply distribution	18	what share the ten largest wallets hold
Trading volume	15	whether real people are trading
Momentum	13	price change over the last 30 minutes
Liquidity	11	real capital sitting in the trading pool
Position on the curve	9	pump.fun only
Community	9	reply counts, Telegram channel
Freshness	9	how long the token has existed
Social presence	6	whether the site and accounts are real
Contract safety	10	honeypot, taxes, mint authority, freeze authority

On top of scoring there is a hard filter: fail one required threshold and the token drops to the Reserve list with the reason attached, however high its score. For example supply concentrated beyond the limit, a contract that still has mint authority, or an unreadable holder table.

Glossary of metrics

Sniper: Wallets that buy within the first seconds of a token's life, usually bots. A high sniper share means most of the supply was taken by a few parties before anyone else knew the token existed.
Bundling: Several buys packed into a single transaction block. It is how one buyer is split across many wallets so the holder distribution looks cleaner than it is.
Same-funder wallets: Distinct wallets whose incoming funds all trace back to one parent wallet. One party counted as many — the scanner traces funding back and groups them.
Developer holdings: The share of supply still held by whoever created the token. The higher it is, the more a single sale can move the price.
Bonding curve (Curve %): pump.fun only: how far the pool has filled before the token graduates to a decentralised exchange. Below 20% is very early; above 95% is near the end of that phase.
Wash trading / market making: Volume produced by one group trading with itself to make the numbers look active. The scanner reads the actual order flow to separate this from real volume.
Honeypot: A contract that lets you buy but not sell, or sets a sell tax high enough that selling is pointless. The scanner checks GoPlus for this on EVM chains.
Locked liquidity: Capital in the trading pool locked for a period, so the creator cannot drain it immediately.
Real-vs-fake score: A 0–100 scale for tokens newly listed on CoinMarketCap: 0 means every signal is verifiable, 100 means most of it is self-reported. Built from domain age, GitHub activity, the gap between self-reported and measured market cap, and the share of Telegram members online.

Data sources and limits

The scanner has no data of its own. It queries public sources and merges them: DexScreener (prices, pairs), GeckoTerminal (new pools, daily candles), RugCheck (holder tables and risk scores on Solana), GoPlus (contract safety on EVM chains), the public pump.fun API, and the PumpPortal websocket for newly created tokens.

Limits worth knowing before you read the table. GeckoTerminal allows roughly 30 requests a minute, so a multi-platform scan takes several minutes and the table always lags the market. Holder data is only available when the source has it; very new tokens usually have none. And every metric measures the recent past — none of them says anything about what happens next.

Frequently asked questions

What does this coin scanner do?: It reads public data on newly listed tokens — price, volume, holder tables, order flow, contract details — then scores each token 0–100 against a fixed set of criteria, and states which metric each token failed on. It does not trade and does not give advice.
Where does the data come from?: DexScreener, GeckoTerminal, RugCheck, GoPlus, the public pump.fun API and the PumpPortal websocket. All open sources; no paid keys.
How often does the table update?: The server scans on a fixed schedule: pump.fun and multi-platform every 5 minutes, newly listed CoinMarketCap tokens every 5 minutes, the mid and high market-cap groups every 15 minutes, and the track-record table every 3 hours.
Does a high score mean it is worth buying?: No. The score only says that at scan time the token met preset thresholds on supply distribution, liquidity and contract safety. It does not forecast price and does not replace your own research. Most new tokens lose nearly all their value regardless of score.
Why is a token that is trending not in the table?: The filter drops tokens there is not enough data to measure — no on-chain pair yet, no readable holder table, or simply too new. Those sit in the Reserve list with the reason they failed, rather than disappearing.
Do I need an account?: No. This page is read-only: no run buttons, no editable thresholds, and everyone sees the same table the server produced.