How long do Pump.fun tokens stay active? Mostly about a day. In a snapshot of 13,219 Pump.fun tokens taken on 14 August 2026, 93% of bonding-curve tokens under a day old had traded more than $1,000 in the previous 24 hours. Among tokens one to two days old the share was 30%, and after a week it was under 9%.
Key findings
- Day one is where almost all of the activity is. Median 24-hour volume on the bonding curve falls from $40,415 for tokens under a day old to $168 for tokens one to two days old.
- Graduating to PumpSwap roughly doubles the share of tokens still trading on day two: 55% against 30% on the curve.
- After two weeks both groups look the same: only 5% to 8% of tokens still clear $1,000 of daily volume.
- Among tokens that are still active, buys and sells are close to even at every age, with median buy shares between 47% and 54%.
- This is one snapshot of a DexScreener-sourced sample, not a census of every Pump.fun launch. The limits are spelled out below.
What did we measure?
PumpWave ran a token tracker from 28 June to 14 August 2026 that collected Pump.fun tokens from two public sources: the DexScreener API feed of the latest token profiles, and DexScreener search results for generic Solana and Pump.fun terms. Only Solana tokens whose address ends in "pump" were kept. On 14 August 2026 at 03:00 UTC every tracked token was refreshed in the same pass, which gives a single, synchronized picture of 13,338 tokens at one moment.
For each token we read the main trading pair (the one with the most liquidity), its venue, the time that pair was created, and its 24-hour volume, buy count, sell count and liquidity. We then grouped tokens by venue - still on the Pump.fun bonding curve, or already in a PumpSwap pool - and by how old the pair was at the moment of the snapshot. We dropped 43 tokens whose main pair was on another venue and 76 with no creation time, which leaves 13,219 tokens: 6,760 on the curve and 6,459 in PumpSwap pools.
A token counts as "active" in this study if it traded at least $1,000 in the 24 hours before the snapshot. The threshold is a choice, so the table further down also gives the $10,000 and $100,000 cut-offs and the median volume for every group.
How fast does activity fall after launch?
On the bonding curve the drop happens in a single step. Of 160 tokens under a day old, 93% cleared $1,000 of volume and 75% cleared $10,000, with a median of $40,415 and 1,127 transactions. One day later the same measures read 30%, 8% and a median of $168 across just 10 transactions. By the third day the median curve token traded $34, and from the first week onward the median is a few dollars.
The long tail is flat rather than falling to zero. Between 7% and 9% of curve tokens older than a week still trade more than $1,000 a day. That share barely moves between one week and a month, which suggests that a token which survives its first few days tends to keep a small, steady audience.
Does graduating to PumpSwap change the curve?
For tokens in a PumpSwap pool, age is measured from the moment the pool was created, which is the moment the bonding curve completed - so these numbers describe life after graduation, not after launch. The first day after migration is the busiest window in the whole dataset: 99% of tokens traded over $1,000, 58% traded over $100,000, and the median was $125,527 across 4,120 transactions.
Graduation buys time, but not much. One to two days after migration 55% of PumpSwap tokens were still active, against 30% of curve tokens of the same age. Two to three days after migration the share was 27%, and between three and seven days it held near 20%. After two weeks the advantage is gone: 5.4% of PumpSwap tokens aged 14 to 30 days were active, slightly below the 7.7% of curve tokens. What happens around the migration itself is covered in the PumpSwap volume bot guide.
Liquidity in the PumpSwap pools that stayed active grows with age, from a median of $2,723 on day one to $12,596 after a month. That is most likely survivorship: pools that kept trading are the ones that kept or attracted liquidity, not a sign that every pool deepens over time.
Are active tokens bought more than they are sold?
No group in the data shows a lopsided tape. Among tokens that were active, the median share of buys in 24-hour transactions sits between 47% and 54% at every age and on both venues. Day-one curve tokens lean slightly toward buys at 53.7%, and tokens one to two days old lean slightly toward sells at 47.4%, which fits a first day of entries followed by a day of exits. Beyond that the split is close to even.
Full results by age and venue
| Venue and age | Tokens | Over $1k | Over $10k | Over $100k | Median 24h volume | Median 24h txns |
|---|---|---|---|---|---|---|
| Curve, under 1 day | 160 | 93.1% | 75.0% | 16.3% | $40,415 | 1,127 |
| Curve, 1-2 days | 206 | 29.6% | 8.3% | 0% | $168 | 10 |
| Curve, 2-3 days | 187 | 11.2% | 3.7% | 0% | $34 | 2 |
| Curve, 3-7 days | 592 | 9.3% | 1.9% | 0% | $5 | 1 |
| Curve, 7-14 days | 1,028 | 8.8% | 1.0% | 0% | $4 | 1 |
| Curve, 14-30 days | 2,033 | 7.7% | 0.7% | 0% | $3 | 1 |
| Curve, 30+ days | 2,554 | 6.6% | 1.3% | 0% | $3 | 1 |
| PumpSwap, under 1 day | 156 | 99.4% | 98.1% | 57.7% | $125,527 | 4,120 |
| PumpSwap, 1-2 days | 200 | 55.0% | 27.0% | 11.0% | $1,435 | 59 |
| PumpSwap, 2-3 days | 186 | 27.4% | 11.8% | 3.8% | $86 | 9 |
| PumpSwap, 3-7 days | 657 | 19.8% | 8.8% | 3.0% | $13 | 3 |
| PumpSwap, 7-14 days | 1,013 | 9.4% | 3.8% | 0.9% | $4 | 2 |
| PumpSwap, 14-30 days | 2,175 | 5.4% | 1.6% | 0.2% | $3 | 1 |
| PumpSwap, 30+ days | 2,072 | 5.6% | 1.6% | 0.6% | $2 | 1 |
What does this mean for timing a launch?
The data describes a market where attention arrives in the first hours and is mostly gone by the second day. That has three practical consequences for anyone planning launch activity, whether that is a community push, paid callers or a Pump.fun volume bot session.
- Be ready before the first hour, not after it. The window in which 93% of curve tokens still trade is the first day. Image, description, links and a first post should exist before anything else happens, because a late start lands in the 30% day.
- Plan for the second day, not only the first. The steepest fall is between day one and day two. A session that covers the first hours and then stops leaves the token exactly where most tokens go quiet; a longer, thinner session that runs into the second day is aimed at that gap. How much volume is enough covers how to size it.
- Treat the migration as a second launch. The day after graduation is the busiest window in the dataset, and the activity fades over the following two to three days. Activity that follows the token into its pool meets that window instead of missing it.
None of this predicts what any single token will do, and no tool can buy a token a place in the 7% that are still trading a month later. What the numbers show is where the attention is, and when. The step-by-step version of a session is in how to run a Pump.fun volume bot session, and what each board actually sorts on is in does a volume bot get you trending.
What are the limits of this data?
- The sample is not every Pump.fun launch. Tokens entered the tracker through the DexScreener token-profile feed and DexScreener search. Projects that pay for a DexScreener profile or that already show up in search are more active than an average launch, so the activity shares here are likely higher than across all Pump.fun tokens.
- It is one moment, not a history. Each token is measured once, at 03:00 UTC on 14 August 2026. The age groups compare different tokens of different ages, not the same tokens over time.
- PumpSwap age starts at migration. For graduated tokens the pair age is time since the pool was created, not since the token was minted.
- Volume is DexScreener volume. Figures follow DexScreener definitions for 24-hour volume and transactions on the main pair only.
- Activity is not success. Clearing $1,000 a day says a token is still traded. It says nothing about who is trading it or whether holders made or lost money.
Can I use this data?
Yes. The aggregated results by venue and age are available as a CSV file: pump-fun-token-activity-2026-08-14.csv. They are published under CC BY 4.0: you can reuse and republish them with a link back to this page as the source, cited as "PumpWave, Pump.fun token activity study, 14 August 2026 snapshot". Questions about the method go to support@pumpwave.net.
Questions people actually ask
What share of Pump.fun tokens are still trading after a week?
In this snapshot, between 6% and 9% of tokens older than a week still traded more than $1,000 in 24 hours: 8.8% of bonding-curve tokens aged 7 to 14 days and 9.4% of PumpSwap tokens 7 to 14 days after migration. After a month the shares were 6.6% and 5.6%.
Do tokens that graduate to PumpSwap stay active longer?
For a few days, yes. One to two days after migration 55% of PumpSwap tokens were still active, against 30% of curve tokens of the same age. After two weeks the difference disappears, with both groups between 5% and 8%.
When is a Pump.fun token most active?
In its first 24 hours on the curve and in the first 24 hours after migration. Day-one curve tokens had a median 24-hour volume of $40,415; day-one PumpSwap pools had $125,527. Both medians fall by more than 90% on the following day.
Where does this data come from?
From the public DexScreener API. Tokens were collected from the DexScreener token-profile feed and search between 28 June and 14 August 2026, and all of them were refreshed in one pass on 14 August 2026 at 03:00 UTC.
Does this mean a volume bot keeps a token alive?
No. The study measures what tokens did, not what caused it, and no tool decides whether people keep trading a token. What the data shows is timing: the attention window is short, and activity that is planned around the first two days and the migration meets it.
Run one and watch it land
Paste a mint, shape the session, see the exact fee before you pay anything. Flat 1% from 100 SOL covers every network and swap cost, no install, no seed phrase.
The people who build the PumpWave session console. We write about Solana launch mechanics from the operator side: what settles on chain, what a session costs and what the board does with it. Corrections and arguments to support@pumpwave.net.