gud tek
Radar for new coinsScore public. Losses too.Paper book since Sep 29, 2026

Radar finds new coin. gud tek draws the line. You decide.

New coin every two seconds on Solana. Nobody can watch that. Machine counts them all and watches the movers.

Radar watches new coins, and older ones that wake up. Coin does something worth a look: Radar pings. gud tek takes pinged coin and draws two lines. Entry and Low. Price breaks Entry: that's the spot. Price loses Low: you were wrong. Both lines known before you're in. Size of loss is not: a stop can fill under the Low.

Score first

1,348paper tickets, $100 each, every chain
294hit 2x (22%)
32peaked at 10x or more (of 1,305 since Oct 4)
−$32,377whole book

Paper book, Sep 29 to Nov 19, 2026, 6:01 AM ET. Since Oct 4: 1,305 tickets, down $31,049. Without the best 3 tickets: down $34,510. The 32 tickets that peaked at 10x or more: book is up $6,392 on them, banked plus still riding.

Most tickets lose. A few run. Book is down. We show it anyway. A radar that hides its losses is a tip sheet.

Why watch a losing book? Staircase pings: 16% reach 2x within 72 hours. Random pick from Live Scout's intake: 2%. More 2x shots. Also more dead coins: 58% of staircase pings dead in 72 hours, random 32%. Staircase has not cleared our own edge bar. Nor has any other lane on our scorecard, pullback, reignite and ⭐ Star among them. Machine finds the shots. Exits are human work.

Lane numbers: 21 days to Nov 19, 2026, scored 72 hours after ping. Staircase n=1,957 pings, typical ping 0.17x. Random pick n=391, typical pick 0.86x.

What it does, what it can't

Does

  • Counts new Solana coins off the launch feed. 316,932 in seven days, Nov 4 to Nov 10.
  • Sorts movers into lanes. One lane, one kind of move.
  • Labels what it sees: ⭐ Star, 📦 bundle, ⚪ contract not read.
  • Since Nov 9, hides any coin GMGN has ever flagged as a honeypot, unless a hand review cleared it. Not on Solana: GMGN gives no honeypot answer there, so a Solana coin on the board is not proof it is safe.
  • Labels bundles and tracks where their coins go. Coins moved to other wallets are not counted as sold.
  • Draws Entry and Low before the break, and saves the time it drew them.

Can't

  • Can't sell for you. Half off at 2x, guard a stop. Nothing more.
  • Can't be first. Measured Oct 7: typical break seen about 51 seconds late, one in ten over 5 minutes late.
  • Can't read every contract. 74% of closed-ticket losses to Nov 10 sat on coins with no security or funder read.
  • Can't promise uptime. Feeds failed silent for weeks, each on its own dates, one bad on most days since July 7. Sweep that found them all ran Oct 9. Fixed, and said out loud.
  • Can't make a losing ticket win.

The method: Fresh Token Hunt

Free guide. Read a ping. Read the labels. Check contract, not ticker. Wait for the line. Size small. Half out at 2x.

Read the guide: The Fresh Token Hunt

Videos come after launch, as often as they can be made and checked. What radar found. What radar gets wrong. How to read a coin with gud tek.

What it costs to run

These will be the real monthly bills for running Radar and gud tek, listed in full. The coin below exists to pay them.

The bill table goes here once every figure comes from a real invoice. No estimated figures go in this table.

The coin

One coin, launching on Pump.fun. Its creator fees will pay the bills above and pay the builder. That is its whole job. It is not an investment and nobody here promises its price will go up. Most tickets in our own paper book lose, and the book is down, as the score above shows.

The coin launches with no bundle, no presale and no hidden wallets. The developer buy will be small, and its wallet address will be published here at launch. The fee wallet will be public.

TO DO on launch day: coin name, ticker, contract address, dev wallet, fee wallet, and Radar's own card for the coin.

Fee ledger

Every dollar of fees in, every bill paid and every payment to the builder is listed here with its transaction or receipt.

DateWhatInOutReceipt
Empty until launch day.

Follow