SmugCalls · live tracker · updated 2026-11-20 11:00:02 UTC

pump.fun graduation rate

1.97%

of pump.fun launches in the week of 9 Nov – 15 Nov 2026 graduated within 24 hours.

SmugCalls counted 218,156 launches in that week; 4,303 migrated off the bonding curve within a day of being created — 1.97%. Counting only the 162,797 launches that traded at least once, the rate is 2.64%. The week before it was 2.03% — down 0.06 points.

How many pump.fun tokens are there?

Since 15 Jun 2026 our collector has recorded 5,350,385 pump.fun launches — about 34,573 a day over the last 4 full weeks — and 129,189 of them have graduated so far (2.41%). These are lower bounds: the collector does not capture every creation, and pump.fun itself has been running since January 2024, long before our feed starts. What happens when a token fills its curve: migration threshold and timing; the curve math: bonding curve calculator.

Week by week

0%1%2%3%Week of 2026-06-15: 0.70%0.715 JunWeek of 2026-06-22: 0.83%0.8Week of 2026-06-29: 1.45%1.429 JunWeek of 2026-07-06: 1.85%1.8Week of 2026-07-13: 2.18%2.213 JulWeek of 2026-07-20: 2.77%2.8Week of 2026-07-27: 2.58%2.627 JulWeek of 2026-08-03: 2.44%2.4Week of 2026-08-10: 2.52%2.510 AugWeek of 2026-08-17: 2.50%2.5Week of 2026-08-24: 2.51%2.524 AugWeek of 2026-08-31: 3.07%3.1Week of 2026-09-07: 3.24%3.27 SepWeek of 2026-09-14: 2.97%3.0Week of 2026-09-21: 2.75%2.821 SepWeek of 2026-09-28: 2.36%2.4Week of 2026-10-05: 2.10%2.15 OctWeek of 2026-10-12: 2.12%2.1Week of 2026-10-19: 1.99%2.019 OctWeek of 2026-10-26: 2.00%2.0Week of 2026-11-02: 2.03%2.02 NovWeek of 2026-11-09: 1.97%2.0Week of 2026-11-16: 1.79%1.816 Novweek starting (Monday, UTC)

finalstill settlingcollector outage that week

The numbers

week (Mon–Sun, UTC)launchestraded
at least once
graduated
within 24h
rate,
all launches
rate,
traded only
graduated
to date
status
16 Nov – 22 Nov102,55977,3951,8401.79%2.38%1.84%in progress
9 Nov – 15 Nov218,156162,7974,3031.97%2.64%2.06%final
2 Nov – 8 Nov230,178166,5524,6742.03%2.81%2.16%final
26 Oct – 1 Nov243,364170,9384,8772.00%2.85%2.12%final
19 Oct – 25 Oct276,355179,0315,5101.99%3.08%2.12%final
12 Oct – 18 Oct268,488187,2085,6972.12%3.04%2.22%final
5 Oct – 11 Oct334,593256,9387,0232.10%2.73%2.20%final
28 Sep – 4 Oct320,670269,9037,5602.36%2.80%2.55%outage
21 Sep – 27 Sep247,639216,1246,8122.75%3.15%2.91%final
14 Sep – 20 Sep232,470192,9996,9072.97%3.58%3.19%final
7 Sep – 13 Sep202,439165,8336,5573.24%3.95%3.56%final
31 Aug – 6 Sep211,438188,0676,4993.07%3.46%3.35%final
24 Aug – 30 Aug310,684272,4467,7872.51%2.86%2.75%final
17 Aug – 23 Aug264,008224,1236,5982.50%2.94%2.74%final
10 Aug – 16 Aug265,531223,6336,6832.52%2.99%2.70%final
3 Aug – 9 Aug248,652214,8016,0612.44%2.82%2.66%final
27 Jul – 2 Aug231,646200,6935,9882.58%2.98%2.90%final
20 Jul – 26 Jul187,665160,3445,2042.77%3.25%3.08%final
13 Jul – 19 Jul163,655144,8363,5752.18%2.47%2.46%final
6 Jul – 12 Jul175,384155,6173,2371.85%2.08%2.10%final
29 Jun – 5 Jul218,218184,1713,1531.45%1.71%1.60%final
22 Jun – 28 Jun185,756144,1961,5440.83%1.07%0.91%final
15 Jun – 21 Jun181,800134,8951,2780.70%0.95%0.81%final

Download the same series: graduation-weekly.json · graduation-weekly.csv. Licensed CC BY 4.0 — use it freely, credit smugcalls.com.

How it is counted

Cohorts, not calendar ratios. A week is the set of tokens created between Monday 00:00 and Sunday 24:00 UTC. Each token is followed individually. Dividing “graduations this week” by “launches this week” mixes tokens of different ages and is not what this page does.

Why 24 hours. Graduation is mostly fast, but not only: in cohorts old enough to judge, about 8% of all graduations arrived later than a day after launch, a few of them months later. A rate without a horizon therefore keeps creeping upward and cannot be quoted. The 24-hour figure is fixed once every token of the week is a day old — it is marked final and does not change again. The “to date” column is the open-ended count, shown for comparison.

Two denominators, both stated. “All launches” counts every token creation seen on-chain, including the ones nobody ever traded. “Traded only” drops those. Published graduation rates differ mostly because this choice goes unstated — the field note walks through it.

What graduation means here. The token migrated off the bonding curve, read from the on-chain migration event rather than inferred from price. 24 graduations in the record carry no usable migration timestamp; they count in “to date” but not in the 24-hour figure. No sampling: the feed starts on 12 June 2026 and every launch it recorded since is in the count.

Weeks still running. Whether a launch ever traded is written to the record about a day after its creation. A week that is not final is therefore counted only over launches older than 26 hours, in every column; the files carry the full count so far as launches_seen. Those rows are provisional.

Known gap. A launch enters the record when the collector captures its creation. An audit on 2 October 2026 found that some creations are not captured, so launch counts are a lower bound and the rates may be shifted. No correction is applied: the size of the gap is being measured.

Week of 28 Sep: Collector was down for 39 and 64 minutes on 29 Sep; launches inside those gaps are missing.

Check the file, not the frog

How many pump.fun launches actually graduate? · questions, answered with numbers · every call on record · bonding curve calculator · migration threshold · about SmugDeg · home