Not "does it eventually leave the bonding curve" — that took hours or days and told you nothing about whether it was worth watching. The question that matters is the one everybody is actually asking when they paste a fresh mint into this site: does it 10x in the first five minutes. So that is what gets measured, every launch, published either way.
What "ran" means. The token's market cap reached 10x what it started at, inside 5 minutes of being minted. Measured straight off pump.fun's own recorded curve reserves and its own all-time-high figure — a fact anyone can re-check, not a model's opinion.
Why the sample size is next to every number. A short perfect record happens by chance. A rate resting on 40 observations is not the same claim as one resting on thousands, so the two are never printed to look alike, and a rate is not called "enough to trust" until enough actual hits have happened, not just enough launches watched.
None of this is a prediction of price. The overwhelming majority of everything measured here goes nowhere at all, in either five minutes or five days. The honest use of this number is calibrating expectations before you paste an address, not a reason to buy anything.
Everything above is measured after the fact: outcomes the tracker was already watching when they happened. This is different — every row below was committed to disk before the outcome existed, using only signals knowable at that moment: swarm notability, the launching wallet's own history, name triggers, and whether the launch's own claimed account is one that's actually mattered before (Elon, Trump and a short, named watchlist — see below). It runs continuously, right now, every ~20 minutes. Nothing here can be edited after the fact, and every miss is counted the same as every hit — that is what makes the number below real.
Coordinated mints — several wallets minting the same name inside a short window — are the loudest thing this system watches for. Whether that pattern says anything about a 5-minute run, rather than about graduation, is a live, ongoing measurement. It is early and it is said plainly below.
The terminal badges a launch when its name references a known trigger figure. This is the receipts behind that badge: every one of those launches, and what actually happened to it, updated live — not asserted once and left to go stale.
Pump.fun records the exact twitter/x.com link a creator pastes in at mint time. That's a sharper key than a name — a link is an account, not a guess — so it's tracked two ways: a short, named watch list of accounts big enough to matter, and every handle that shows up on more than one launch, regardless of who it is. No live tweets are read to build this; every input here is a field pump.fun already hands back for a token this site was asked to check anyway.
coincheck