We keep a record of pump.fun launches and check back on them. Not the winners, not the ones that trend. Just whatever came through the feed, tracked on a timer so the numbers are not chosen after the fact.
This is the first cut of that record: 381 launches, each followed for at least six hours after it was created. Below is what the median coin did, in numbers.
The median pump.fun coin does not move
Market cap at each checkpoint, median across the 381 coins:
| Time after launch | Median market cap |
|---|---|
| 1 minute | $2,833 |
| 5 minutes | $2,799 |
| 15 minutes | $2,782 |
| 30 minutes | $2,780 |
| 1 hour | $2,772 |
| 3 hours | $2,772 |
| 6 hours | $2,768 |
Read the first row against the last. Two thousand eight hundred thirty three dollars to two thousand seven hundred sixty eight. The median launch loses about two percent over six hours and never gets off the floor. If you opened any of these coins an hour later you would see the same chart you saw at minute one, only flatter.
Half of them peak in the first five minutes
For every coin we recorded, we noted when its highest market cap of the six hours happened. 196 of 381, or 51 percent, peaked within five minutes of launch. Whatever that peak was, the rest of the window was spent below it.
That is the shape of the problem. The first five minutes are the most expensive window on the whole chart, and it is also the window most people are looking at, because that is when the coin is moving and when it appears in feeds.
The ladder to 100,000 is very thin
| Peak reached | Coins | Share |
|---|---|---|
| $5,000 | 79 of 381 | 21% |
| $10,000 | 45 of 381 | 12% |
| $25,000 | 26 of 381 | 7% |
| $50,000 | 17 of 381 | 4% |
| $100,000 | 10 of 381 | 3% |
| $500,000 | 6 of 381 | 2% |
Three in a hundred reached one hundred thousand dollars at any point. Two in a hundred reached half a million. The rest of the feed is the 97 percent that never printed a number worth screenshotting.
Thirty nine percent were still above their first minute print an hour later
We compared each coin’s one hour market cap with its one minute market cap. The median ratio was 1.00, meaning the median coin was exactly where it started. Only 39 percent were above their minute one level an hour in. The other 61 percent were below it.
That is a low bar and most coins still fail it. An hour of holding a random launch is, on this sample, a coin flip that pays nothing.
What we do with this
This is why the board at firememecoins.com ranks the ten coins with the most live buying rather than the newest or the loudest, and why every card shows its own grade instead of a promise. Freshness is free on pump.fun. Buying is not.
If you want the check on a single coin instead of a list, the checker grades any mint on supply, bundling, exit cost and buying pressure, and tells you what a five hundred dollar exit would do to the price before you size anything. If you are new to how the curve actually works, this explains graduation in plain terms.
Method
Sample: 381 launches recorded by our own collector, each with at least four follow up price samples across six hours. Market cap figures are medians, taken from the latest sample at or before each checkpoint, and rounded. The 837 launches in our record include coins with shorter observation windows, which is why the ladder above is drawn only from the 381 with a full six hour trace. Nothing here is a forecast. It is what the record says, and the record keeps growing.
