pump.fun graduation rate: daily launches and graduations
How many new coins launch on pump.fun each day, how many of them graduate to PumpSwap within 24 hours, how fast, and how many still trade a month later. Counted by PumpFYI from a public on-chain event stream since 4 November 2026 and checked against the chain every day.
Collecting since 2026-11-04 19:08 UTC; first full-day figures on 2026-11-06, and the first graduation rate (the share of a day's launches that graduate within 24 hours) once 3 full days of launches have had 24 hours each, expected 2026-11-09. So far PumpFYI has counted 402 new pump.fun coins and 18 graduations to PumpSwap in 25 minutes of observation, to 2026-11-04 19:37 UTC.
What this means: A coin "graduates" when buyers fill its bonding curve, at which point it moves to its own PumpSwap pool (pump.fun charges 0.015 SOL for this). Most coins never get there: published estimates run from about 0.6% to 3.1%, depending on the month and on whether bot-driven launches are removed. Our headline follows each day's launches for 24 hours and counts how many graduate; coins we did not see launch are left out. Graduating is not success: a graduated coin can still lose most of its value, which is why we also track whether graduates still trade 30 days later (the first full-day cohort reaches 30 days on 2026-12-05).
- Collecting since
- 2026-11-04
- 19:08 UTC
- Launches counted
- 402
- 0.4 hours observed
- Graduations counted
- 18
- moved to PumpSwap
- Stream completeness
- first check Nov 5
- vs the chain, daily
- First graduation rate
- 2026-11-09
- after 3 full-day cohorts
Counts only. The daily chart starts once there are three days of data; hatched bars are hours not fully observed. Collecting since 2026-11-04 19:08 UTC.
Day by day
UTC days, newest first. "Graduated within 24h" follows the coins launched that day and fills in 24 hours after the day ends; "still trading at 30d" fills in after 30 days. "Seen vs chain" is the daily completeness check.
| Date (UTC) | Launches | Graduations | Graduated within 24h | Excl. bundles | Per launch, same day | Seen vs chain | Still trading at 30d |
|---|---|---|---|---|---|---|---|
| 2026-11-04 partial, 2% | 402 | 18 | pending | pending | full days only | pending | pending |
Showing all 1 days.Full table as CSV
Why published graduation rates disagree
Every figure below measures something slightly different: the window, whether bots are filtered, and whether it follows a cohort of coins or the daily flow.
| Source | Rate | Window |
|---|---|---|
| PumpFYI (this page) | pending | from 2026-11-09 |
| PumpFunData | 0.84% | April 2026 |
| madeonsol.comas reported by CoinPaprika | 1.76% | August 2026 |
| Marino et al. (arXiv 2602.14860)as reported by CoinPaprika | 0.63% | 1 Sep to 1 Oct 2025 |
A cohort rate follows one day's launches to see how many graduate. A same-day ratio compares a day's graduations (of coins launched at any time) with that day's launches, so it runs higher when older coins are graduating. Our 24-hour figure understates the share that ever graduate, because a few coins graduate days later; the 7-day figure in the CSV covers those.
How this is measured
Counts of on-chain pump.fun events observed through PumpPortal's public websocket (wss://pumpportal.fun/api/data, subscribeNewToken + subscribeMigration), a third-party feed not run by pump.fun. creates = new coins launched on pump.fun's bonding curve (pool 'pump'); creates_other_pools = launches the feed reports for other launchpads (e.g. bonk), excluded from rates. graduations = migration events (a coin completing its bonding curve and moving to its PumpSwap pool) seen that UTC day, whenever the coin was created. grad_ratio = graduations / creates on the same day (a flow ratio, not a probability). ttg = time from creation to graduation, only for graduating coins whose creation we also observed (ttg_n), so early in the series it leans toward fast graduations. grads_under_1m = timed graduations less than 60 seconds after creation, almost always a launch bundled with buys that fill its own curve. cohort_grad_24h / _7d = share of coins created that day that graduated within 24 hours / 7 days of creation (null until the window has passed). survival_30d = share of a day's graduates that, 30 days later, still have a Solana pair on DexScreener with at least $1,000 of liquidity and non-zero 24h volume. coverage = share of the day's 1,440 minutes our socket was connected for the whole minute; partial = coverage below 98% or the day is still running, so counts on partial days undercount. creates_observed / graduations_observed = counts inside those fully observed minutes, used for per-day rates. cohort_complete = the launch day and the following day were both at least 98% observed, so the cohort's 24-hour window is fully covered. cohort_grad_24h_ex_bundles = the 24h cohort rate with coins that graduated less than 60 seconds after launch removed from both sides. crosscheck = a daily completeness check against the chain: successful transactions on pump.fun's migration account (public Solana RPC) versus the migrations the stream reported; missed_share is the estimated share of on-chain migrations the stream did not report. PumpPortal's terms (section 6) restrict republishing its feed, so PumpFYI publishes only daily aggregate counts of public on-chain events; per-coin records stay on our server. The series starts 2026-11-04 at 19:08 UTC: there is no earlier data from this collector.
- Series start. Counting began on 2026-11-04 19:08 UTC. There is no earlier PumpFYI data. The headline rate appears only once 3 full days of launches have each been followed for 24 hours; until then this page shows counts, not a rate.
- The headline is a cohort rate. We take the coins launched on one UTC day and count how many graduate within 24 hours of their own launch. A migration of a coin whose launch we did not see is left out of every cohort. The "per launch, same day" column is a different, flow measure and is shown only for full days.
- Bundled launches. Some coins graduate within seconds because the launch transaction is bundled with buys that fill the whole curve. We report the cohort rate with and without coins that graduated less than 60 seconds after launch.
- Completeness. Each day records how many of its 1,440 minutes our connection was up. Each completed day is also checked against the chain: transactions on pump.fun's migration account (public Solana RPC) are compared, coin by coin, with the graduations the stream reported, and a sample of the unmatched ones is inspected. If the stream missed more than 10% of graduations on a recent day, this page says so.
- What we publish. The event stream is PumpPortal's free public feed, a third party not affiliated with pump.fun. PumpFYI publishes only daily aggregate counts of public on-chain events; per-coin records stay on our server. pump.fun itself publishes no daily launch or graduation count.
Source: PumpFYI collector on PumpPortal's public websocket (new-token and migration events), checked daily against the chain. Snapshot taken 2026-11-04 19:37 UTC. PumpFYI is independent and unofficial: not pump.fun, and not endorsed by it. We never use pump.fun's Developer API for datasets. Not investment advice: most memecoins lose most of their value.
Questions people ask
What is the pump.fun graduation rate?
Published studies put the share of pump.fun coins that graduate at roughly 0.6% to 3.1%, depending on the month and on whether bot-driven launches are removed. PumpFYI's own daily figure starts once three full days of launches have been followed for 24 hours, expected 2026-11-09.
How many coins are launched on pump.fun per day?
PumpFYI's first full-day count publishes on 2026-11-06. Earlier studies counted about 25,000 a day in April 2026 and about 33,000 a day in August 2026.
What happens when a pump.fun coin graduates?
Its bonding curve fills, trading on the curve stops, and the coin moves to its own PumpSwap pool, where the fee steps down as its market cap grows. pump.fun charges a 0.015 SOL graduation fee. See how graduation works.
Does graduating mean a coin is safe or successful?
No. Graduation only means buyers filled the curve; the coin can still be dumped by its creator or early holders and lose most of its value. Check a coin before you buy it, and read how to spot a rug.
Cite or reuse this data
Free to reuse with attribution under CC BY 4.0. Machine-readable: /data/graduation.json and /data/graduation.csv.
Suggested citation:
PumpFYI (2026). pump.fun graduation rate: daily launches and graduations. https://pumpfyi.com/data/graduation/ (data as of 2026-11-04 19:37 UTC; source: PumpFYI count of PumpPortal's public event stream, checked against the chain).