Bots Pile Into 1 in 4 New pump.fun Tokens at Launch, and Almost None Graduate

Pump.fun remains the busiest token factory in crypto, and DEXTools sampled what actually happens to its tokens in the first minutes of life. Across 215 freshly launched pump.fun tokens checked on June 29, 2026, the data splits into two extremes: a minority get hammered by automated trading the instant they launch, while most barely trade at all. Nearly a quarter showed a bot-like burst of activity within 15 minutes, 30 percent were already dead with zero trades, the median token logged just 3 trades, and not a single one in the sample had graduated to Raydium at the time of the check.

What a bot launch burst looks like

On a permissionless launchpad, automated traders compete to be first into anything that might run. When a token suddenly logs dozens of trades within minutes of creation, far above the handful a normal launch sees, that is the signature of bots sniping the launch rather than organic demand. In this sample, 24 percent of new pump.fun tokens crossed that threshold, with the busiest tokens (the 90th percentile) logging 54 trades in the window. A burst is not automatically bullish; it often just means snipers are racing each other in and back out, and a retail buyer arriving seconds later is buying their exit.

Most new tokens barely trade

The flip side is the silent majority. The median freshly launched token had just 3 trades, 30 percent recorded no trades at all, and only 5 percent showed steady, moving volume rather than a single blip. Median pooled liquidity sat around 4,400 dollars, the baseline the bonding curve seeds, but liquidity alone does not make a market when nobody is trading.

Graduation is the rare exception

On pump.fun, a token graduates when it completes its bonding curve and its liquidity migrates to Raydium, the milestone that separates a fleeting experiment from something with a real market. In this snapshot, zero of the 215 sampled tokens had graduated at the time of the check. That is consistent with the wider pattern across chains, where only a tiny fraction of launches build any lasting market, and it is why graduation rate is one of the most important filters when sorting signal from noise on a launchpad.

We track this live in the Pump.fun Bot Activity Index, which updates the burst, dead and graduation shares continuously.

How to protect yourself

The takeaway for anyone trading new launches is that the first-minute tape is mostly machines and dead weight, not opportunity. If a token shows an instant burst, assume bots are already positioned and that you may be exit liquidity. If it shows nothing, there is no market to speak of. Either way, the responsible move is to check before you buy: confirm real and locked liquidity, look for sustained rather than spiked volume, and scan the contract for red flags. Our Token Safety Checker automates the safety scan, and our guides on buying Solana memecoins safely and the anti-rug playbook cover the workflow. If you want to understand the launchpad mechanics themselves, see how memecoins are created.

Methodology and disclaimer: figures are a live snapshot from the DEXTools Pump.fun Bot Activity Index as of June 29, 2026, based on 215 sampled pools via the DEXTools API. Numbers change continuously as new tokens launch. Bot-like is an activity heuristic, not a definitive identification of automated wallets. This article is for information only and is not financial advice.